Case Study: Použitý obsah populárnych článkov z Hacker News

http://kukuruku.co/hub/diy/usb-killer

USB devices are ubiquitous tools in modern technology, but not all of them are harmless. Among the most concerning developments is the “USB Killer”—a device specifically designed to damage or destroy hardware through a high-voltage electrical surge. This guide explores the mechanics of USB Killers, the risks they pose, and practical steps to safeguard your devices and sensitive data from these malicious tools. What Is a USB Killer?

A USB Killer is a seemingly harmless device that resembles a standard USB drive. However, its purpose is anything but benign. When connected to a USB port, it delivers a high-voltage surge back into the system, often resulting in irreparable damage to the device. Originally developed as a proof-of-concept for identifying vulnerabilities, USB Killers are now being misused as malicious tools. How Does It Work?

Energy Harvesting: The device extracts power from the USB port and stores it in internal capacitors. Voltage Surge: It rapidly discharges the stored energy, creating a voltage surge that overwhelms the system. Hardware Damage: Key components like the motherboard, USB controllers, and other connected peripherals can be permanently damaged.

Understanding the Risks

The potential damage caused by USB Killers goes beyond physical hardware destruction. These tools can lead to:

Data Loss: Essential files and information may become irretrievable. Operational Downtime: Damaged systems can disrupt workflows in businesses or personal use. Financial Costs: Repairs or replacements for affected devices can be costly.

How to Protect Your Devices from USB Killers

Avoid Untrusted USB Devices Only use USB devices from reputable sources. Be cautious of promotional USB drives or devices with unknown origins. Implement Physical Security Measures Use port blockers or locks to restrict access to USB ports on your devices. Store critical devices in secure locations to prevent unauthorized access. Disable Unused Ports Disable unused USB ports in your system’s BIOS or through administrative controls. Limit USB access to only essential devices. Invest in USB Protection Devices Specialized USB protection tools, such as data-only cables, can prevent electrical surges. Endpoint security software can monitor USB activities and block suspicious devices.

Broader Implications for Data Integrity

While USB Killers represent a clear threat to physical devices, they also highlight the importance of protecting data and content from less visible risks. Just as malicious hardware can compromise systems, unverified or plagiarized content can undermine credibility in digital spaces.

For example, ensuring the originality of your content is crucial in academic, professional, and creative contexts. Tools like Paper-Checker.com provide advanced solutions for detecting plagiarism and AI-generated content. By integrating such tools into your workflow, you can maintain the integrity of your work and protect intellectual property. Conclusion

USB Killers serve as a stark reminder of the vulnerabilities inherent in modern technology. By understanding how these devices operate and adopting preventive measures, you can safeguard your hardware and data. Similarly, taking steps to protect the integrity of your digital content ensures credibility and trustworthiness.

Whether securing your devices from malicious tools or ensuring originality in your content, proactive measures are key to staying safe in a rapidly evolving technological landscape.

---

https://www.zdnet.com/article/former-student-destroys-59-university-computers-using-usb-killer-device/

An Indian national in the US has pleaded guilty this week to destroying 59 computers at the College of St. Rose, in New York, using a weaponized USB thumb drive named "USB Killer" that he purchased online. Security

How passkeys work: Your passwordless journey begins here This infamous people search site is back after leaking 3 billion records - how to remove your data from it ASAP Syncable vs. non-syncable passkeys: Are roaming authenticators the best of both worlds? Yes, you need a firewall on Linux - here's why and which to use

The incident took place on February 14, according to court documents obtained by ZDNet, and the suspect, Vishwanath Akuthota, 27, filmed himself while destroying some of the computers.

"I'm going to kill this guy," "it's dead," and "it's gone. Boom," Akuthota said on recordings obtained by the prosecution.

The suspect destroyed 59 computers, but also seven computer monitors and computer-enhanced podiums that had open USB slots.

He did it using USB Killer, a weaponized thumb drive that he purchased from a well-known online store that sells these types of devices.

USB Killer devices work by rapidly charging thumb drive capacitors from the USB power supply, and then discharging the electrical current back into the USB slot --all in a matter of seconds-- effectively frying the computer to which the USB Killer device is connected.

Equipment damages amounted to $51,109, along with $7,362 in employee time for investigating and replacing destroyed hardware, which Akuthota agreed to pay as part of the plea deal, according to court documents.

He graduated College of St. Rose in 2017 with an MBA, and on the night of the incident, he was no longer a student at the college and was residing in the US on a student visa.

Akuthota was arrested on February 22 and will be sentenced later this year, on August 12. He faces up to ten years in prison, a fine of up to $250,000, and a term of post-imprisonment supervised release of up to 3 years.

---

https://arstechnica.com/gadgets/2016/12/usb-killer-fries-devices/

Last year we wrote about the “USB Killer”—a DIY USB stick that fried almost everything (laptops, smartphones, consoles, cars) that it was plugged into. Now the USB Killer has been mass produced—you can buy it online for about £50/$50. Now everyone can destroy just about every computer that has a USB port. Hooray.

The commercialised USB Killer looks like a fairly humdrum memory stick. You can even purchase a “Test Shield” for £15/$15, which lets you try out the kill stick—watch the spark of electricity arc between the two wires!—without actually frying the target device, though I’m not sure why you would want to spend £65 to do that. The website proudly states that the USB Killer is CE approved, meaning it has passed a number of EU electrical safety directives. The original USB Killer prototype, more clearly showing you the capacitors (the square components) and other parts of the system. USB Killer version 1.0, with hand-soldered bits.

The USB Killer is shockingly simple in its operation. As soon as you plug it in, a DC-to-DC converter starts drawing power from the host system and storing electricity in its bank of capacitors (the square-shaped components). When the capacitors reach a potential of -220V, the device dumps all of that electricity into the USB data lines, most likely frying whatever is on the other end. If the host doesn’t just roll over and die, the USB stick does the charge-discharge process again and again until it sizzles.

Since the USB Killer has gone on sale, it has been used to fry laptops (including an old ThinkPad and a brand new MacBook Pro), an Xbox One, the new Google Pixel phone, and some cars (infotainment units, rather than whole cars… for now). Notably, some devices fare better than others, and there’s a range of possible outcomes—the USB Killer doesn’t just nuke everything completely.

In the video below you can see that a Galaxy Note 7 loses its USB port (so it can’t be charged), but otherwise remains functional; likewise, a new iPad Pro freaks out while the stick is plugged in, but seems to regain consciousness afterwards. Curiously, the Pixel is fine when a third-party USB-C converter is used, but using the official Google dongle results in a dead device. One guy fries a ton of stuff with the USB Killer.

In another video it seems the iPhone 7 Plus suffers a similar fate to the Note 7: the Lightning port is fried, but the rest of the device is okay. You’ll also be glad to hear that the iPhone 4 and iPhone 3GS, connected via a 30-pin dongle, both seem to be immune to the USB Killer. All told, the company behind USB Killer says that 95 percent of devices are susceptible to a USB power surge attack.

Without taking a fried device apart it’s impossible to say how extensive the damage actually is. When you see the screen go black, and then not return after a reboot attempt, it’s likely that the surge travelled to the CPU or some other core component. If the victim device was a desktop PC, you might get away with replacing the motherboard—on a mobile or embedded device, diagnosing and fixing the issue is probably going to be more effort than it’s worth.

A better solution, of course, is protecting a system against the USB Killer in the first place, though that isn’t an easy task. Electrically, the most simple solution is an opto-isolator: a chip that uses an LED paired with a photodiode to physically isolate one electrical circuit from another. As devices move towards USB-C there’s another possible solution: USB authentication.

Neither of those solutions help protect the hundreds of millions—perhaps billions—of devices in the world with unprotected USB ports, though. Cars, airplanes, routers, machines that control industrial centrifuges… in those cases, the only real defence is physically capping ports or educating people to never insert unknown hardware.

---

https://www.usbkill.com/

USBKill V4.0 The new USBKill needs no host power. |

The most powerful USBKill ever. New unstoppable attack modes. Remote controlled. The ultimate pentesting device. More Powerful

Stronger discharge, unlimited run time, bypasses all USB-C & Lightning security. New Attack Modes

Smart-phone, BLE Remote Control, Time-delay, Hidden Magnet Trigger and Classic. Test Everything

15+ Adaptors, including: USB-C, iPhone, VGA, DisplayPort, HDMI, MicroUSB & more. Never miss a target

Internal Rechargable battery means the USBKill V4 even on devices that are switched off. What is the USBKill? ↷

The USBKill is a device that stress tests hardware. When plugged in power is taken from a USB-Port, multiplied, and discharged into the data-lines, typically disabling an unprotected device.

See the USBKill in action ⚡

Watch the V4 Pro tested against a 2020 Macbook Pro, iPhone 11, Samsung S20, HDMI TV, DisplayPort, Network Device and more..

Used by penetration testers, hardware manufacturers, law-enforcement and industrial clients world-wide, the USBKill has been adopted the industry standard for USB Stress-testing.

The USBKill V4 caters specifically to industry needs: significantly more powerful, more flexible, more covert and more compatible.

Meet the USBKill V4

The USBKill V4 brings automation, remote control, improved discretion, enhanced performance over a vast vulnerability surface.

Learn more now. USBKill V4 Feature Overview ⚡ Powerful Hardware Update

The USBKill has evolved beyond a simple plug-and-zap device. The new V4 hardware framework enables advanced functionality and performance.

Aside from more powerful discharges and improved stability, the V4 has an internal, rechargeable battery, which allows for "Offline Attacks" - where the host device is not turned on.

"Offline Mode" also bypasses all known USB-C and Lightning (Apple/iPhone) security protocols, rendering the V4 the ultimate device for testing smartphones and modern hardware.

⏱️ Advanced Attack Modes

The USBKill has evolved, and so have the ways to use it. V4 Introduces single pulse and continuous pulse attack modes. By default, the V4 will not activate until triggered - giving pentesters and LEA ultimate discretion. The V4 can be triggered in multiple ways:

Remote Trigger: Trigger a single or continuous attack via dedicated remote control, up to 100m away from the device. Smartphone Trigger: Control and trigger single or continuous attacks via the included Android and iOS smartphone app. Timed Attack: Schedule a date and time when the USBKill will trigger. The V4 can stay dormant without power for over 200 days. Magnetic Trigger: Trigger the USBKill with a magnet (stylish and covert magnetic ring included) by passing your hand over the device. Classic Mode: Triggers the instant it is plugged into a USB drive.

🔌 Exhaustive Accessories

The USBKill V4 boasts extensive accessories and adaptors, preparing you for any situation.

USBKill V4 Tester: The V4 Tester has been completely redesigned to function as a multi-function shield device, providing high-voltage protection and juice-jacking protection.

The V4 Adaptors are broken into two familes, Basic and Advanced.

Basic Adaptors: USB-C, Lighting (iPhone), MicroUSB, MiniUSB, USB-A Female.

Advanced Adaptors: USB-B Male, USB-B Female, VGA (DB15), HDMI Male, HDMI Female-Female, DisplayPort Male, DisplayPort Female-Female, RJ45 Male, RJ-45 Female

The V4 accessories open a vast attack surface for pentesters, covering all possible types of hardware: smartphones, computers, laptops, printers, televisions, network equipment, USB drives, external harddrives, and much more.

USBKill V4 Pro

The Professional version of the V4 has wireless, remote & smartphone control, internal battery for offline attacks, is compatible with all accessories, and can be configured in Classic Mode (plug-and-zap) mode.

Pro: Kit → Pro: Standalone →

USBKill V4 Basic

The Basic Edition of the V4 has an internal battery for offline attacks and is compatible with all accessories. It has no wireless and must be triggered manually via the covert Magnet Trigger Ring.

Basic: Kit → Basic: Standalone →

Classic Edition

The Classic Edition is the original USBKill device, with the updated V4 stability. Low-cost, requires no configuration or charging - simply "Plug and Zap".

Classic: Standalone →

Tactical Pack

Designed for Penetration Testers and Industrial clients, the Tactical Pack has every device, accessory and adaptor available, in a professional tool case.

Tactical Kit →

---

https://www.techspot.com/news/108394-usb-killer-dead-apple-drops-firewire-support-macos.html

The "USB killer" is dead: Apple drops FireWire support in macOS 26 The final nail in the coffin for FireWire By Alfonso Maruccia June 20, 2025 at 12:13 PM 14 comments The USB killer is dead: Apple drops FireWire support in macOS 26 Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust.

What just happened? FireWire was Apple's ambitious attempt to establish a novel connectivity technology for its computer ecosystem back in the 90s. Now, Cupertino is unceremoniously removing the standard from its modern operating system capabilities.

Recently introduced macOS 26 "Tahoe" will bring a new naming scheme for Apple's operating system ecosystem, novel features, and a highly questionable Vista-like UI redesign. The upcoming release will also mark the end of official support for outdated FireWire hardware. The so-called "USB killer" that was ultimately defeated by its own irrelevance is no more, at least according to developers testing the beta versions of the new OS.

macOS 26 beta is no longer compatible with first-generation iPods, or likely any other FireWire-based devices. The technology is being fully deprecated, reports confirm. In Tahoe, System Utilities and Finder no longer display FireWire storage devices, even when they are connected via Thunderbolt-to-FireWire adapters.

Officially known as IEEE 1394, the FireWire standard was developed by Apple in collaboration with other tech companies, including Sony and Panasonic. Apple first introduced the serial bus technology in some Macintosh models in 1999, as a replacement for the parallel SCSI bus. FireWire competed with USB at a time when it could achieve data transfer rates up to 30 times faster.

Despite offering more hardware features and better performance, FireWire was largely confined to Apple's computing ecosystem. The technology was popular for digital video editing, audio equipment, and other professional tasks, but Apple eventually scrapped it as USB and other competing technologies such as Thunderbolt grew in popularity and capability.

Cupertino shipped its last iMac with a native FireWire port in 2011, but macOS continued supporting IEEE 1394 technology for several more years. Now, macOS 26 will kill support for good, forcing owners of older hardware peripherals and archivists to preserve legacy Macs or avoid updating the OS entirely. // Related Stories

Performance and design complaints mount after iOS 26's Liquid Glass launch Apple's new MacBook Pro 14 M5 is faster and better, but déjà vu

Apple routinely phases out older technology support in macOS, as the company isn't particularly interested in preserving backward compatibility in the way Microsoft largely does with Windows. However, macOS 26 will be especially brutal in this regard, as it will also be the last version to support Intel-based Macs.

---

http://hj.diva-portal.org/smash/record.jsf?pid=diva2%3A1683816&dswid=1160

Analysis of effective energy consumption of Bluetooth Low Energy versus Bluetooth Classic Tåqvist, Carl Jönköping University, School of Engineering, JTH, Department of Computing. Luks, Jonathan Jönköping University, School of Engineering, JTH, Department of Computing. 2022 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesis Abstract [en]

Wireless technology is used daily across the globe. A very common wireless technology is Bluetooth. The Bluetooth technology exists everywhere, from cars to mobile phones and even kitchen appliances. Recently, Bluetooth Low Energy has added support for another physical layer, LE 2M PHY. This physical layer is supposed to be faster and more energy efficient than its predecessor, LE 1M PHY, with a decrease in range. Because of this new physical layer, Bluetooth Low Energy can now compete with Bluetooth Classic during data transmission, in both speed and energy efficiency. This thesis aims to find the breaking point where Bluetooth Low Energy becomes less energy efficient than Bluetooth Classic, in relation to bit rate speed and total amount of bytes sent. Before experiments were conducted, multiple iterations of an artifact had to be done to end up with an artifact that provides valid and reliable data. The experiments were then conducted by changing the bit rate speed and sending different amounts of bytes. The results from the experiments show that Bluetooth Classic is practically both faster and more energy efficient with its fastest modulation than Bluetooth Low Energy is with LE 2M PHY enabled, even though this should not be the case theoretically. Bluetooth Classic is overall more energy efficient than Bluetooth Low Energy and thus the conclusion of this study is that no breaking points between the two technologies exist. Place, publisher, year, edition, pages 2022. , p. 45 Keywords [en] Bluetooth Classic, Bluetooth Low Energy, Energy efficiency, Internet of Things, Throughput, Wireless technologies National Category Energy Systems Identifiers URN: urn:nbn:se:hj:diva-57970ISRN: JU-JTH-DTA-1-20220170OAI: oai:DiVA.org:hj-57970DiVA, id: diva2:1683816 External cooperation Knowit Subject / course JTH, Computer Engineering Supervisors Axelsson, Andreas Examiners Adlemo, Anders Available from: 2022-08-04 Created: 2022-07-18 Last updated: 2025-10-13Bibliographically approved

Analysis of effective energy consumption of Bluetooth Low Energy versus Bluetooth Classic Main Subject area: Computer Engineering Author: Jonathan Luks, Carl Tåqvist Supervisor: Andreas Axelsson JÖNKÖPING 2022 June i This final thesis has been carried out at the School of Engineering at Jönköping University within Computer Engineering. The authors are responsible for the presented opinions, conclusions, and results. Examiner: Anders Adlemo Supervisor: Andreas Axelsson Scope: 15 hp (First-cycle education) Date: 2022-06-01 ii Abstract Wireless technology is used daily across the globe. A very common wireless technology is Bluetooth. The Bluetooth technology exists everywhere, from cars to mobile phones and even kitchen appliances. Recently, Bluetooth Low Energy has added support for another physical layer, LE 2M PHY. This physical layer is supposed to be faster and more energy efficient than its predecessor, LE 1M PHY, with a decrease in range. Because of this new physical layer, Bluetooth Low Energy can now compete with Bluetooth Classic during data transmission, in both speed and energy efficiency. This thesis aims to find the breaking point where Bluetooth Low Energy becomes less energy efficient than Bluetooth Classic, in relation to bit rate speed and total amount of bytes sent. Before experiments were conducted, multiple iterations of an artifact had to be done to end up with an artifact that provides valid and reliable data. The experiments were then conducted by changing the bit rate speed and sending different amounts of bytes. The results from the experiments show that Bluetooth Classic is practically both faster and more energy efficient with its fastest modulation than Bluetooth Low Energy is with LE 2M PHY enabled, even though this should not be the case theoretically. Bluetooth Classic is overall more energy efficient than Bluetooth Low Energy and thus the conclusion of this study is that no breaking points between the two technologies exist. Keywords Bluetooth Classic, Bluetooth Low Energy, Energy efficiency, Internet of Things, Throughput, Wireless technologies. iii Acknowledgement We would like to give a huge thanks to Filip Doversten from Knowit for all the help we got with the technical aspect of our thesis work. A huge thanks should also be given to Knowit for letting us do our thesis work with them and for helping us with equipment and guidance. We would also like to thank Andreas Axelsson, our supervisor from Jönköping University, for the guidance we got through our thesis work. iv Table of content Abstract .......................................................................................... ii Keywords ........................................................................................ ii Acknowledgement ........................................................................ iii Table of content ............................................................................ iv

1. Introduction ................................................................................ 6 1.1 PROBLEM STATEMENT ........................................................................................... 6 1.2 PURPOSE AND RESEARCH QUESTIONS .................................................................... 7 1.3 SCOPE AND LIMITATIONS....................................................................................... 7 1.4 DISPOSITION .......................................................................................................... 8
2. Method and implementation .................................................... 9 2.1 DATA COLLECTION ................................................................................................ 9 2.1.1 Design Science Research .............................................................................. 9 2.1.2 Experiments ................................................................................................ 13 2.2 DATA ANALYSIS .................................................................................................. 14 2.3 VALIDITY AND RELIABILITY ................................................................................ 14 2.4 CONSIDERATIONS ................................................................................................ 15
3. Technical framework ............................................................... 16 3.1 BLUETOOTH ........................................................................................................ 16 3.1.1 Host ............................................................................................................. 16 3.1.2 Controller .................................................................................................... 18 3.1.3 Host Controller Interface ............................................................................ 22 3.1.4 Bluetooth Serial Port Profile ....................................................................... 22 3.2 ELECTRICAL CONCEPTS ....................................................................................... 23 3.2.1 Small voltages ............................................................................................. 24 3.2.2 Burden voltage ............................................................................................ 24 3.2.3 Shunt resistors ............................................................................................. 24 3.2.4 Brownout..................................................................................................... 25 3.3 WHY BLUETOOTH LOW ENERGY IS MORE ENERGY EFFICIENT ............................. 25 v
4. Results ....................................................................................... 26 4.1 EMPIRICAL DATA ................................................................................................. 26
5. Analysis ..................................................................................... 29 5.1 ESTIMATED THROUGHPUT VERSUS MEASURED THROUGHPUT .............................. 29 5.2 ANALYSIS OF RESEARCH QUESTIONS ................................................................... 29 5.2.1 Research question 1 .................................................................................... 29 5.2.2 Research question 2 .................................................................................... 31 5.2.3 Research question 3 .................................................................................... 34
6. Discussion ................................................................................. 35 6.1 RESULT DISCUSSION ............................................................................................ 35 6.1.1 Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on bit rate speed? ...................................................................................................... 35 6.1.2 Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on total amount of data sent? .................................................................................... 36 6.1.3 How can a circuit be built to accurately measure the required data to validate and verify the two earlier research questions? ..................................................... 36 6.2 METHOD DISCUSSION .......................................................................................... 37
7. Conclusions and further research........................................... 38 7.1 CONCLUSIONS ..................................................................................................... 38 7.1.1 Implications................................................................................................. 38 7.2 FURTHER RESEARCH............................................................................................ 38
8. References ................................................................................ 39
9. Appendix ................................................................................... 41 6
10. Introduction 1.1 Problem statement Most people use wireless technology every day. Wireless technology could for example be a tv remote control or wireless headphones. Though both being wireless, they use diverse types of communication technology. Remote controls most commonly use infrared communication (Maker.io Staff, 2022) while wireless headsets use Bluetooth as a connector between the wireless object and the object to which it is connected to. The focus in this thesis is Bluetooth technology, which is described more closely further on. Bluetooth is the leading technology in wireless data communication and the global standard for simple and secure wireless connections (Bluetooth, 2022). Between devices with Bluetooth technology, data can be transferred, for example music between a phone and earbuds. Bluetooth can be found in most devices which supports pairing with other devices. Bluetooth´s aim was to offer a general-purpose connectivity solution (Marcel, 2018). Over time, however, Bluetooth was primarily used for audio streaming. To return to the original idea of offering the general purpose of Bluetooth technology, Bluetooth Low Energy (BLE) was developed after the first Bluetooth Classic or as it is called Basic Rate / Enhanced Data Rate (BR/EDR). Whereas BR/EDR offers a higher data transfer speed than BLE, it is more suitable for audio streaming and data transfers. BLE, that does not yet support audio streaming, supports data transfers, location services and device networks (Bluetooth, 2022). Since BLE offers device positioning it can track presence, distance, and direction to other devices. BLE also offers three types of communication, namely point-to-point, broadcast, and mesh. With the Bluetooth 5.0 release, BLE got a bit rate speed upgrade to support 2Mb/s, which is only 1Mb/s slower compared to BR/EDR. This opened new possibilities for BLE and was later announced that audio streaming would become available. This makes it directly compete with the advantage of BR/EDR that were the only Bluetooth variant to support it. For small devices with a low battery capacity this could be a good addition to be able to save battery. Earlier studies have been done about the efficiency of BLE, for example (Bulić, Kojek, & Biasizzo, 2019) concluded that the number of bytes sent was the variable that affected the energy consumption. Their study was conducted by only measuring BLE efficiency and did not compare it to BR/EDR. According to (Bergelin & Ericsson, 2019) there might be a breaking point where BLE is less efficient than BR/EDR and thus, this study gives a better insight about the efficiency of BLE. This might also make the decision whether to change from BR/EDR to BLE for developers easier and better explain the difference for the customers. And when looking up energy consumption for BLE and BR/EDR in data sheets there is only shown the power consumption of one modulation for each technology and not comparing the different modulations. Other studies have not researched in the area about the breaking point and only about the efficiency of BLE or other similar alternatives such as Zigbee (Siekkinen, Hiienkari, K. Nurminen, & Nieminen, 2012). 7 Developers now a days has two different Bluetooth communication protocols to choose from. Companies, such as Knowit (Knowit, 2022), who work with Bluetooth modules for automotives would like to know if and where this breaking point could be. The breaking point is the point where BLE becomes less efficient than BR/EDR when it comes to the energy consumption. 1.2 Purpose and research questions The purpose of this study is to find the breaking point where Bluetooth Classic becomes more energy efficient than Bluetooth Low Energy when active. To achieve this, the data of the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy in relation to bit rate speed and total amount of data sent must be measured and compared. To measure this data a circuit must be created with enough reliability to provide data with low deviation. According to (Bergelin & Ericsson, 2019), there could be a breaking point where Bluetooth Classic and Bluetooth Low Energy cross in their energy consumption efficiency. (Bergelin & Ericsson, 2019) did not research this area and therefore leads to this study’s first research question:

• Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on bit rate speed? Related to the first research question, if a breaking point exists when measuring efficiency on bit rate speed, then there must also be a breaking point when measuring efficiency related to the total amount of data sent between Bluetooth devices. Thus, this study’s second research question is as followed:
• Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on total amount of data sent? To validate the two previous research questions, a circuit must be built. The circuit must be precise and accurate to measure valid data with the least amount of disturbance from external sources. This leads to this study’s third research question:
• How can a circuit be built to accurately measure the required data to validate and verify the two earlier research questions? 1.3 Scope and limitations Since this study is a continuation on another study’s suggested research (Bergelin & Ericsson, 2019), the experiments must be conducted in a comparable way. The experiments made by Bergelin, and Ericsson (2019) were conducted with the use of an Android application supporting version 8.0 and newer. This study does not use the previous study’s developed Generic Attribute Profile (GATT), but instead use a Bluetooth stack provided by Knowit, called blueGO 8 (Knowit, 2022) The blueGO stack is built upon another Bluetooth stack made by Opensynergy, called Blue SDK, which uses their own developed GATT. The blueGO stack is only supported on a limited number of Bluetooth modules and thus the experiments are limited to those specific modules. When measuring with the artifact, the base current is excluded from the energy consumption used by the Bluetooth module to get more accurate results. The previous study also conducts experiments with external disturbance added, which this study does not entail. As this thesis work is conducted at Knowit a direction more towards automotive was chosen. This reflects the hardware chosen which the experiments were conducted with as the Jody-W263 is an automotive graded Bluetooth and Wi-Fi module. This is also the reason why the BLE physical layer LE Coded PHY were not measured with in any experiment. LE Coded PHY is a BLE modulation with a much better range for low power wireless communication technology. As this physical layer is mostly seen in use cases where greater range could be an advantage, the smart home sector is one example of it. The greater range is then something that is not of use when only connecting for example a mobile phone or earphones that will always stay within a short range inside of an automotive. 1.4 Disposition This study is divided into six chapters, including this chapter of introduction. The chapter “Method and implementation” describes how this study’s process of data collection and analysis is laid out. The next chapter, “Technical framework”, explains the fundamental knowledge that is needed to conduct similar experiments. The chapter after is the “Results” chapter, where the empirical data is displayed and presented. The analysed results and methods used in this study are discussed in the chapter “Analysis”. The next chapter is “Discussion” where the analysed data and the methods used are discussed. The study is then concluded in the last chapter, “Conclusions and further research”. 9

2. Method and implementation For this study, Design Science Research (DSR) (vom Brocke, Hevner, & Maedche,

2020. is the best methodology to use when building something that must be reliable and provide accurate results. DSR develops artifacts in many iterations in to get products that meets specified requirements. Hence, by using this methodology when creating a circuit, a reliable, accurate and complete circuit has been made and therefore, valid, and reliable data could be collected. Figure 1: Example of a DSR iteration To answer this study’s question 1 and 2 there will be a collection of empirical data from the artifact created for this study. To then answer the questions a regression analysis will be made to find the relations. 2.1 Data collection The purpose of this study can only be fulfilled by recording data and thus, empirical research is the method of choice to collect the data. The data was recorded in real-time using the developed artifact. The data collected was the current going through the Bluetooth modules chip. The data was collected by iterating through different bit rate speeds and then within these speed settings iterate through various numbers of bytes sent. 2.1.1 Design Science Research This subchapter explains the different versions of the artifact that were developed by using DSR. 2.1.1.1 Hardware Artifact The artifact was iterated and built around the Jody-W164 Bluetooth module, using different hardware and software to find the most reliable and accurate artifact. The first iteration of the artifact was a simple circuit consisting of a power supply, an ampere meter, the Jody-W164 Bluetooth module, a resistor, and a laptop. The laptop was always connected to the Bluetooth module to initialize and setup the Bluetooth module. 10 Figure 2: Circuit Diagram of first DSR Iteration. As seen in Figure 2, the Bluetooth module, resistor, power supply and ampere meter are connected in series. The passive voltage probe from the oscilloscope was then connected to the circuit across the resistor to read the voltage across said resistor. This iteration provided readable data but having the ampere meter connected in series with the rest of the circuit could add burden voltage, which can provide unreliable data if not accounted for. A USB cable connected to the Bluetooth module was needed for data communication. Figure 3: Circuit Diagram of second DSR Iteration. 11 The next iteration, as displayed in Figure 3, is similar to the first iteration. The only difference is the probe used. The second iteration of the circuit used a Current Probe, which instead of measuring directly on the circuit, it measures the magnetic field around the wire it is clamped on to. The second iteration of the artifact provided similar data, but it was difficult to read and evaluate as the data the current probe provided had a lot of noise. Figure 4: Circuit Diagram of third DSR Iteration. The third iteration of the artifact used the Power Profiler Kit 2 (PPK2) manufactured by Nordic Semiconductor instead of a regular circuit. The PPK2 acts as a power source to the Bluetooth module and is also able to measure the current being used by the module. The PPK2 has software that is also made by Nordic Semiconductor, which looks like a simplified oscilloscope. The software is needed to run the PPK2 properly and it gives a variety measurement such as average current used, and total electric charge used. The software also displays the electric signal as a graph, just like a regular oscilloscope. The PPK2 can supply 500mA and up to 1A, for 1A supply a second USB cable needs to be connected. For the configuration in Figure 4 with only one cable connected to the Bluetooth module it was possible to both power it and read correct measurements. 12 Figure 5: Circuit Diagram of fourth DSR Iteration The fourth iteration of the artifact were made to update the Bluetooth modules that both supported the Bluetooth 5.0 core specification. The Bluetooth modules also had to support the optional feature of 2 Mb/s for BLE, the Bluetooth module chosen were a Jody-W263 as seen in Figure 5 and the USB Bluetooth dongle changed to an Asus USB-BT500. As the Bluetooth dongle does not connect to the hardware artifact it was decided to not be part of the figure. Some changes had to be made with the hardware artifact as the Jody-W263 could not be supplied through the PPK2. By cutting open the USB cable that will supply the Jody-W263, the PPK2 could be connected in serial to still be able to measure the Bluetooth module. The PPK2 still had to be connected with another USB cable to get power itself and data communication with the computer. To isolate current that is not from anything else other than the Bluetooth chip the jumper for the LED´s were removed and only having Bluetooth activated. 13 2.1.1.2 Software Artifact Figure 6: Flowchart diagram for software artifact. The software side of the artifact were made up of Knowit´s Bluetooth stack BlueGO. With their help and knowledge, a software artifact was made that made it possible to send bytes ranging from 0x00 to 0xFF. Both technologies (BLE and BR/EDR) were made to be as closely implemented as possible. From Figure 6 a demonstration of how the artifact work is shown. One thing to mention is that with BLE the last packet sent were not the correct data but sent an uncontrolled data packet. After discussions and with the time available for this thesis project it was not investigated further to try and fix this since there should not be any difference for the collected data. The correct amount of data was sent, and all other packets had the correct data sent. This problem was not met with BR/EDR. The software artifact was only implemented on the transmitting Bluetooth module since the measured data was taken from the receiving Bluetooth module it was not needed. 2.1.2 Experiments Experiments to evaluate the artifacts were conducted with a Ublox Jody-W164 module that supports Bluetooth 4.6 and a USB Bluetooth module with support of Bluetooth 4.0. This was done due to the simple reason that it was the one available at the time and similar enough. For the artifact evaluating it was decided to only use Bluetooth Classic for the purpose of not having to reinitialize the Bluetooth modules and set everything up again. For the testing of the artifacts three different amounts of bytes were used, 50k, 500k and 1M bytes. The reason for only using three diffirent numbers of bytes were 14 only because of the simplicity of testing the different artifacts. Later when doing the experiments with the final iteration it was decided not to use 1M bytes but instead use lower amounts of bytes between 50k and 500k. The reason why 1M were not used in the final experiments were because of the extreme amount of bytes and such amounts rarely appears in real life applications. This gave a better insight into the gap between 50k and 500k and a fourth and fifth test were added. For the final experiments it was decided to use 50k, 100k, 250k, 350k and 500k. The Ublox Jody-W164 functioned as the receiver while the USB Bluetooth module functioned as a transmitter. Table 1: Table of average measured current through all iterations of the artifacts Number of bytes sent 50k 500k 1M Artifact iteration 1 Average mA: 12.9 12.5 10.6 Artifact iteration 2 Average mA: 11.4 11.3 11.1 Artifact iteration 3 Average mA: 12.42 12.09 12.35 The data collected for Table 1 can be seen under chapter Appendix from figure 24 to

32. The data collection was also collected multiple times to ensure that the collected data was reliable. Each amount of bytes sent were tested three times and then taken an average of. This is done with the version 1 to 3 of the artifact, which is shown in the table. 2.2 Data analysis The data analysis for this study was done by conducting experiments and further observe the collected data. A correlation between the energy consumption and the bit rate speed of both Bluetooth technologies may be found by observing the collected data. And to further investigate if there is a relation between the number of bytes that get sent and its energy consumption. This data is then compared to other studies. 2.3 Validity and reliability To increase the validity and reliability of an academic thesis, it is important to measure relevant data and make sure that the tests and experiments can be replicated correctly. To increase the trustworthiness of this academic thesis, the data for each relevant parameter will be measured multiple times to ensure low deviation in the data. To ensure validity the artifact created for the study are going through iterations of improvements to ensure that correct and relevant data is measured. This is done through the method DSR described earlier with aid from Knowit. 15 To increase the reliability of the study everything is being described thoroughly to make sure the experiments can be remade. All the relevant flowcharts about the artifact and all its iterations are available from the study and easy to replicate. The components are mentioned from the study, and therefore the artifact will be able to be replicated with the same hardware. The change of hardware chosen is described and what advantage it gave to the study. The blueGO stack is licensed and therefore not available for the public to use unless you have a license for it. There should not be any difference when recreating the data collection with an open-source Bluetooth stack. As blueGO is used across many different industries, for example automotives and medical/health devices it should cover more or just as much as what other open-source Bluetooth stacks could have. 2.4 Considerations A consideration to take is that no experiments with artificially created disturbance were conducted in this study. The study tried to minimize the external disturbances as much as possible by measuring both Bluetooth types in the same environments with similar conditions. The experiments were also conducted multiple times to remove the risk of corrupted data and to determine whether the data can be reproduced. It is not possible to eliminate all the external disturbances that could have possibly changed the outcome of the data. 16
33. Technical framework This chapter describes the fundamental knowledge that is needed to understand and follow the experiments for this study and the knowledge needed to conduct similar experiments. 3.1 Bluetooth Bluetooth is a wireless communication technology that uses short-range radio transmissions. The technology is used to transfer data between fixed and mobile devices. The data is transferred using Ultra High Frequency radio waves over the unlicensed Industrial, Scientific, and Medical (ISM) frequency band that can range from 2.4 to 2.48 GHz. Bluetooth can be split into two technologies, BR/EDR and BLE. Bergelin and Ericsson (2019) mentions that there are two stacks which are mutually used in both technologies, they both use a host and a controller stack. The stacks are not equally built because they involve different protocols to offer different solutions. Figure 7: Bluetooth stacks (Tosi, Taffoni, Santacatterina, Sannino, & Formica, 2017) There are three different configurations to be made, see Figure 7. The three stacks are one for BLE, BR/EDR and one for dual mode that supports both technologies. The stacks consist of three parts, host, host controller interface (HCI), and controller. The third option of having a Bluetooth stack that supports both technologies is a combination of all components from respective Bluetooth stacks. 3.1.1 Host To understand the different parts of the host in a bluetooth stack the parts for BR/EDR is explained first and BLE after. BR/EDR: 17

• SDP: The first component Service Discovery Protocol (SDP) is used by a client device to find the services that it can use with the server device. The SDP server maintains a database with services that an application can discover to see what is available and to determine the characteristics of the available services.
• RFCOMM: The Radio Frequency Communication (RFCOMM) protocol is an emulation of RS-232 serial port communication between devices wirelessly.
• L2CAP: The Logical Link and Control Adaptation Protocol (L2CAP) provides a connection- oriented and connectionless data service to the upper layer protocols with protocol multiplexing capability and segmentation and reassembly operations. It permits higher level protocols and applications to transmit and receive upper layer data packets with a size of up to 64 kilobytes in length. L2CAP also permits per-channel flow control and retransmission. The L2CAP layer also provides logical channels, named L2CAP channels, which are multiplexed over at least one or more logical links. BLE:
• GAP: The Generic Access Profile (GAP) has three main purposes: To introduce definitions, recommendations and requirements related to modes and access procedures that will be used by transport and application profiles. To describe how devices should behave in standby and in connecting states to prevent situations where links and channels cannot be established between Bluetooth devices or prevent multi-profile operations. To express requirements for user interface aspects, mainly towards coding schemes and names for procedures and parameters.
• GATT: The GATT defines the service framework by using the Attribute protocol (ATT). It defines all procedures and formats of services. The procedures defined are for example the discovering, writing, reading, notifying, and indicating characteristics. 18
• ATT: ATT allows the device referred to as the server to expose attributes and the associated values to the client. The attributes exposed by the server can be discovered, read, and written by the client and the server can respond, indicate, and notify to the client.
• SMP: The Security Manager Protocol (SMP) is a protocol for the generating of encryption keys and identity keys for peer-to-peer connections. This block also handles the storing of encryption keys and identity keys and are responsible for creating random addresses to know device identities. It will also provide the controller with stored keys during the encryption or pairing procedures.
• L2CAP The LE stack also consists of L2CAP just as BR/EDR has in its stack too. 3.1.2 Controller To understand the different parts of the controller in a bluetooth stack the parts for BR/EDR is explained first and BLE after. BR/EDR:
• LM: Link Manager (LM) is used to control and negotiate operations of Bluetooth connections between two devices. This includes the set-up and control of logical transports, logical links, and physical links. The LM communicates between two devices Link Managers. All LM messages should only use physical links, associated logical links and logical transports between the sending device and the receiving device.
• BR/EDR PHY: BR/EDR radio (physical layer) operates in the 2.4 GHz unlicensed ISM frequency band. To avoid interference and fading it uses a frequency hopper. The frequency hopper is called Adaptive Frequency Hopping (AFH) and is something that both Bluetooth technologies uses. AFH works such as the Bluetooth channels gets divided into two categories, used and unused. The used channels are a part of the hopping sequence and the unused will be replacing the used in a hopping sequence in a pseudo-random order. AFH allows the Bluetooth device to use all its available channels, so for BR/EDR it can use all the 79 channels available but not less than 20. BR/EDR channels range between 2.400 GHz to 2.4835 GHz and all channels are spaced by 1 MHz. The formula to find all BR/EDR operating frequency bands: 𝑓 = 2402 + 𝑘 𝑀𝐻𝑧, 𝑘 = 0, … ,78 19 There are two modes to define the radio, the non-optional mode called Basic Rate (BR) which uses a shaped, binary FM modulation. The second mode which is the optional is called Enhanced Data Rate (EDR). This mode uses PSK modulation and has two variants: π/4-DQPSK and 8DPSK. The symbol rate for BR/EDR is 1 megasymbol per second. The bitrate speed for BR mode is:
• 1 Mb/s And the bitrate speed for the optional mode EDR active:
• 2 Mb/s
• 3 Mb/s The different types of modulations have different package formats. From Figure 8 and Figure 9 the difference between a BR packet and EDR packet is shown. Figure 8: BR packet format The different length of the contents of a BR packet: Access code: 68 or 72 bits. Header: 54 bits. Payload: 0 to 2790 bits. Figure 9: EDR packet format The different length of the contents of a EDR packet: Access code: 68 or 72 bits. Header: 54 bits. Guard: is a defined time period between 4.75 to 5.25 μs. Sync: 20 to 30 bits. EDR payload: 0 to 2790 bits. Trailer: 4 to 6 bits. BR/EDR as mentioned in Introduction only supports one type of communication Topology. BR/EDR supports Point-to-Point communication and are therefore only able to communicate with one device at a time. BLE:
• LL: Link Layer (LL) consists of multiple states which it operates in: 20 Figure 10: State diagram of LL state machine (Bluetooth SIG, Inc., 2016) According to Bluetooth SIG, Inc. (2016) the states are defined as followed: Standby state, in this state the Bluetooth device does not transmit or receive any packets and can be entered from any state. From Figure 10 this state can be entered from any state. Advertising state, in this state the device will be transmitting advertising physical channel packets and is able to listen to and responding to responses triggered by advertising physical channel packets. From Figure 10 this state can only be entered from standby state. Scanning state, in this state the device will be listening for advertising physical channel packets from devices who are advertising. From Figure 10 this state can be entered from standby state. Initiating state, in this state the device is only listening to advertising physical channel packets from a specific device and respond to these packets to initiate a connection with that device. From Figure 10 this state can be entered from standby state. Connection state, in this state the device has a connection to another device. From Figure 10 this state can be entered from Advertising state or Initiating state.
• LE PHY: Just like BR/EDR, BLE also operates in the 2.4 GHz unlicensed ISM frequency band. BLE also uses the frequency-hopping spread spectrum called AFH to avoid interference and fading. BLE has a range between 2.400 GHz to 2.483.5 and 40 channels with a spacing of 2 MHz. The formula to find all BLE operating frequency bands: 𝑓 = 2402 + 2𝑘 𝑀ℎ𝑧, 𝑘 = 0, … , 39 21 BLE has two modulations defined, the first one is the mandatory LE 1M PHY with a symbol rate of 1 Msym/s. The second optional modulation mode is LE 2M PHY with a symbol rate of 2 Msym/s. The bitrate speed of LE 1M PHY is:
• LE 1M PHY, uncoded data, 1 Mb/s
• LE Coded PHY S=2, 500 kb/s
• LE Coded PHY S=8, 125 kb/s And the bitrate speed of LE 2M PHY is:
• LE 2M, uncoded data, 2 Mb/s A connection between BLE two devices differs from BR/EDR communication. BLE have something called Connection Interval. A connection interval is a predefined time to which the host should send data packets with the client until the next packet should be sent. The time which the host exchange data packets with the client is called a Connection Event. When the packets are exchanged and there still is time left of the Connection Interval, the radio goes idle. Figure 11 shows what a Connection Interval looks like. Figure 11: Connection Interval BLE (Tosi, Taffoni, Santacatterina, Sannino, & Formica, 2017) The different types of modulations have different package formats. From Figure 12 and Figure 13 the difference between an uncoded and coded packet are shown. Figure 12: LL packet format LE Uncoded PHY The different length of the contents of a LL packet: Preamble: 8 to 16 bits. Access Address: 32 bits. PDU: 16 to 2056 bits. CRC: 24 bits. Figure 13: LL packet format LE Coded PHY (Bluetooth SIG, Inc., 2016) 22 The different length of the contents of a LL coded packet: Preamble: Uncoded. Access Address: 32 bits. CI: 3 bits. TERM1: 3 bits. PDU: 16 to 2056 bits. CRC: 24 bits. TERM2: 3 bits. As mentioned before in the Introduction, BLE supports different types of connections. BLE supports different types of communication topologies, it supports Point-to-Point, Broadcast and Mesh (Bluetooth, 2022). Broadcasting makes the BLE device able to send packages out to any scanning or receiving device that is in the range of the device. This is the fastest way to send out packets to multiple peers at a time. GAP defines two roles in broadcasting, a broadcaster and observer. The observer continuously scans at periodic intervals to see if there are any packets to receive from the broadcaster. The broadcaster periodically sends out the packets to any device that can receive them. Mesh is another way for BLE to communicate with. In this communication the device acts as both host and client simultaneously. Mesh is an extensive network of multiple devices without any connection but stays within the same network to communicate with each other. And Point-to-Point connection is where connections are established one-to-one (Bluetooth, 2022). 3.1.3 Host Controller Interface The Host Controller Interface manages the connection between the host and the controller of every Bluetooth device, it acts as an agent. The HCI provides a uniform command method for the Host to access Controller capabilities and to control connections to other Controllers. The HCI stores all its commands sent and received and this data can be parsed by doing a so-called “HCI dump”. 3.1.4 Bluetooth Serial Port Profile Bluetooth Serial Port Profile (SPP) is a GAP protocol that is pre-defined within BR/EDR. SPP is a profile that is made to emulate physical serial ports such as RS232 with a virtual serial port through Bluetooth (Bluetooth, 2012). 23 Figure 14: Layers of SPP within GAP (Bluetooth, 2012) The SPP protocol has several sub-protocols that depend on it as shown in Figure 14. Figure 15: Model of SPP (Bluetooth, 2012) The RFCOMM layer within the model displayed by Figure 15 is the layer enabling the wireless communication by emulating a serial port. The point of this emulation is to enable transmitting and receiving data through wireless communication. The amount of data transmitted can vary between a couple of bytes up to multiple large files. 3.2 Electrical concepts To accurately measure the current going through the Bluetooth module, several unique factors must be considered. The biggest factors are explained in this sub-chapter and are as followed: Small Voltages, Burden Voltage, Shunt Resistors, and Brownout. 24 3.2.1 Small voltages To measure electrical current or voltage, a Digital Multimeter (DMM) is very commonly used as they are cheap and easy to set up and measure with. DMMs commonly have resolutions of one millivolt, which means that the DMM can only measure changes at millivolt level. Thus, when measuring voltages lower than one millivolt, a DMM is not a viable tool to measure the voltage. When measuring small voltages, great care needs to be taken, due to low resistances and thus too high current source can damage the electrical device. Other factors such as various noise sources or thermoelectric effects can cause errors or offsets when measuring the voltage (Daire, 2005). Daire (2005) mentions that thermoelectric effects are often the dominant source to voltage offsets when measuring the electrical device and suggests using a “three-point current reversal technique” to minimize the thermally induced voltage. 3.2.2 Burden voltage Burden voltage is the loss of voltage when measuring current through a measuring device. This voltage drop can affect the circuit and thus show incorrect measurement data. The circuit must be well thought out to minimize the amount of burden voltage that may occur while measuring the Bluetooth module. The effect of Burden Voltage can be calculated using Ohm’s law: 𝐼 = 𝑉 𝑅 Figure 16: Circuit examples (NI, 2018) In Figure 16, circuit (a) has a 1.5-volt source with a 5 ohm’s load. Circuit (b) is similar but with a 0.5-volt Burden Voltage due to the added DMM. By using the example circuits, the effects of Burden Voltage can be calculated as such: a) 1.5𝑉 5Ω = 0.3𝐴 b) 1.5𝑉−0.5𝑉 5Ω = 0.2𝐴 These calculations prove how much a DMM can influence the current going through a circuit and how this can effectively corrupt the data if not measured properly. 3.2.3 Shunt resistors According to EEPower (u.d.), a shunt resistor is defined as followed: “A shunt resistor is used to measure electric current, alternating or direct. This is done by measuring the 25 voltage drop across the resistor.” A shunt resistor is commonly used when the electric current going through a circuit is too high for the ampere meter or DMM. By adding a shunt resistor to the circuit, the current is divided between the ammeter and shunt resistor and thus larger currents can be measured (EEPower, n.d.). 3.2.4 Brownout Brownout is a term within electronics that is defined as: “an intentional or unintentional drop in voltage in an electrical power supply system.” This effect can last for minutes up to multiple hours and could potentially damage multiple electrical devices (CD & Power, n.d.). Different to a Blackout, where the electrical power supply system will shut down, current will still flow within the system during a brownout which can possibly cause damage if the brownout occurs spontaneously (Direct Energy, n.d.). After a brownout has occurred, a power surge is bound to happen, which is the biggest factor to malfunctioning electronic devices. 3.3 Why Bluetooth Low Energy is more energy efficient As BLE´s name suggest it is a more energy sufficient technology compared to BR/EDR. But how does BLE manage to do it? The main difference between them is that BLE can have the radio switched on for a short period of time (Donovan, 2011). When data is to be sent connections can be established quickly and will be released as soon as the data has been sent which will be minimizing the power consumption. BLE can scan for other devices, connect, transmit data, confirm valid reception, and then terminate the link in a period as little as 3ms. For BR/EDR this will typically require a period of over hundreds of milliseconds. As BLE only uses three advertising channels the scanning for other devices will be done much quicker than BR/EDR that uses 32 channels. This results in a scan period of only 0.6 to 1.2ms rather than BR/EDR that takes around 22.5ms, this is also one way that BLE can consume less power. BLE also have the advantage of using shorter packets, this can be seen in Figure 12 and Figure 13 for BLE, Figure 8 and Figure 9 for BR/EDR with the amount of bits sent described. The last thing that helps BLE with its efficiency is by using Gaussian frequency shift keying (GFSK). Both BLE and BR/EDR uses GFSK, but the difference is that BR/EDR uses a modulation index 0.35. Meanwhile the modulation for BLE is set somewhere between 0.45 to 0.55 which is closer to the level for Gaussian minimum shift keying. This increase of modulation results in lower power consumed but is more prone to inter- symbol interference. 26

4. Results This chapter presents and analyses the data that were collected during the conducted experiments. 4.1 Empirical data The following tables present the collected data from the conducted experiments. The columns are divided into the total amount of data sent and the rows describe the time it took to send the data, and the average electrical current used during that period. The tables also have a third row, which subtracts the average electrical current used during the module’s idle period from the average current used during the data sending. The average idle for BR/EDR was 35.15 mA, and 33.67 mA for BLE in a connected state and not sending any data. When measuring the current idle the Bluetooth module were set as the host and in connected state with an interval of three seconds. The exact same steps were taken for both Bluetooth technologies. The three next tables are of collected data using Bluetooth Classic with its three different speeds, 1 Mb/s, 2Mb/s, and 3Mb/s. Table 2: Collected data using BR/EDR with 1 Mb/s. Iteration #1 50k 100k 250k 350k 500k Time (s) 3,855 7,705 19,25 26,94 38,49 Average (mA) 39,1 39,06 39,07 39,1 39,1 Idle removed (mA) 3,95 3,91 3,92 3,95 3,95 Iteration #2 50k 100k 250k 350k 500k Time (s) 3,855 7,706 19,24 26,96 38,49 Average (mA) 39,12 39,08 39,08 39,09 39,07 Idle removed (mA) 3,97 3,93 3,93 3,94 3,92 Iteration #3 50k 100k 250k 350k 500k Time (s) 3,857 7,704 19,25 26,97 38,53 Average (mA) 39,09 39,07 39,1 39,07 39,1 Idle removed (mA) 3,94 3,92 3,95 3,92 3,95 Average Time 3,85566667 7,705 19,2466667 26,9566667 38,5033333 Average Effective 3,95333333 3,92 3,93333333 3,93666667 3,94 27 Table 3: Collected data using BR/EDR with 2 Mb/s. Iteration #1 50k 100k 250k 350k 500k Time (s) 0,4916 1,031 2,555 3,587 5,133 Average (mA) 39,47 39,42 39,45 39,44 39,49 Idle removed (mA) 4,32 4,27 4,3 4,29 4,34 Iteration #2 50k 100k 250k 350k 500k Time (s) 0,4931 1,035 2,611 3,558 5,102 Average (mA) 39,48 39,38 39,4 39,47 39,47 Idle removed (mA) 4,33 4,23 4,25 4,32 4,32 Iteration #3 50k 100k 250k 350k 500k Time (s) 0,4952 1,022 2,574 3,549 5,083 Average (mA) 39,42 39,34 39,38 39,48 39,5 Idle removed (mA) 4,27 4,19 4,23 4,33 4,35 Average Time 0,4933 1,02933333 2,58 3,56466667 5,106 Average Effective 4,30666667 4,23 4,26 4,31333333 4,33666667 Table 4: Collected data using BR/EDR with 3 Mb/s. Iteration #1 50k 100k 250k 350k 500k Time (s) 0,3686 0,759 1,911 2,706 3,847 Average (mA) 40,11 40,05 39,99 40,02 40,03 Idle removed (mA) 4,96 4,9 4,84 4,87 4,88 Iteration #2 50k 100k 250k 350k 500k Time (s) 0,3762 0,7674 1,941 2,692 3,867 Average (mA) 40,02 39,95 39,98 40,02 40,03 Idle removed (mA) 4,87 4,8 4,83 4,87 4,88 Iteration #3 50k 100k 250k 350k 500k Time (s) 0,3753 0,7486 1,908 2,726 3,875 Average (mA) 40,14 40,02 39,98 39,98 40,08 Idle removed (mA) 4,99 4,87 4,83 4,83 4,93 Average Time 0,37336667 0,75833333 1,92 2,708 3,863 Average Effective 4,94 4,85666667 4,83333333 4,85666667 4,89666667 The two next tables are of collected data using Bluetooth Low Energy with its two different speeds, 1 Mb/s and 2Mb/s. 28 Table 5: Collected data using BLE with LE 1M PHY. Iteration #1 50k 100k 250k 350k 500k Time (s) 0,7631 1,21 4,027 5,385 7,681 Average (mA) 40,67 40,59 40,74 40,72 40,9 Idle removed (mA) 7 6,92 7,07 7,05 7,23 Iteration #2 50k 100k 250k 350k 500k Time (s) 0,7686 1,687 3,731 5,401 7,594 Average (mA) 40,68 40,29 40,58 40,75 40,76 Idle removed (mA) 7,01 6,62 6,91 7,08 7,09 Iteration #3 50k 100k 250k 350k 500k Time (s) 0,8004 1,539 3,846 5,419 7,769 Average (mA) 40,4 40,55 40,83 40,44 40,64 Idle removed (mA) 6,73 6,88 7,16 6,77 6,97 Average Time 0,77736667 1,47866667 3,868 5,40166667 7,68133333 Average Effective 6,91333333 6,80666667 7,04666667 6,96666667 7,09666667 Table 6: Collected data using BLE with LE 2M PHY. #1 50k 100k 250k 350k 500k Time (s) 0,3948 0,8281 2,177 2,683 4,228 Average (mA) 40,63 40,41 40,01 40,89 40,15 Idle removed (mA) 6,96 6,74 6,34 7,22 6,48 #2 50k 100k 250k 350k 500k Time (s) 0,5 0,9029 2,026 3,046 4,547 Average (mA) 39,19 39,72 40,43 39,97 39,67 Idle removed (mA) 5,52 6,05 6,76 6,3 6 #3 50k 100k 250k 350k 500k Time (s) 0,5373 0,753 1,977 2,753 4,178 Average (mA) 38,92 40,83 40,6 40,61 40,24 Idle removed (mA) 5,25 7,16 6,93 6,94 6,57 Average Time 0,47736667 0,828 2,06 2,82733333 4,31766667 Average Effective 5,91 6,65 6,67666667 6,82 6,35 29
5. Analysis 5.1 Estimated throughput versus measured throughput With the Bluetooth 5.0 release, BLE got a new mode which was advertised as being able to reach up to 2 Mb/s. This speed is only hypothetical and not possible to reach practically. The estimated throughput for BLE according to the Bluetooth SIG’s mathematical formula is around 1.4 Mb/s when using the LE 2M PHY activated (Ren, 2017). The estimated throughput for BR/EDR is 2.1 Mb/s when using the packet type 3-DH5 (Donovan, 2011). The result from this study shows that BR/EDR has a faster throughput than BLE when using its 3 Mb/s modulation. The experiments showed that when using BR/EDR’s 2 Mb/s, the practical throughput were on average 0.786 Mb/s. And while using its 3 Mb/s modulation, the throughput were around 1.047 Mb/s. Meanwhile, when using BLE’s 1M PHY, the average throughput were around 0.522 Mb/s and 0.938 Mb/s when using 2M PHY. The bit rate speed for the BR/EDR 1Mb/s mode were at an average of 0.104 Mb/s. Using these results, it shows that on average, the BR/EDR throughput speeds are faster than the BLE throughput speeds. Table 7: Average measured throughput. Technology BLE BR/EDR Mode 1M PHY 2M PHY 1 Mb/s 2 Mb/s 3 Mb/s Measured Speed (Mb/s) 0.5223 0.9383 0.1038 0.7864 1.0474 The energy consumption efficiency is calculated by multiplying the average measured electrical current during the period the Bluetooth module is receiving data, times the number of seconds that period lasted. This will give the total coulomb units used during that period. ∁ = 𝐴 ∗ 𝑠 Equation 1: Calculation of coulomb 5.2 Analysis of research questions The purpose of analysing the collected data is to answer the study’s research questions. To analyse the data according to the research questions, experimentation and observation and also comparison to other studies and datasheets must be made. 5.2.1 Research question 1

• Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on bit rate speed? 30 5.2.1.1 BLE Figure 17: BLE Total mC used When comparing the different bit rate speeds of BLE looking at Figure 17 a trend can be seen where the LE 2M PHY modulation for BLE tends to show a lower consumed amount of coulomb when receiving data. Furthermore, when looking at Table 5 and Table 6 the real difference is that LE 2M PHY has a lower average of mA drawn under receiving compared to LE 1M PHY. The different modulation makes a difference then by making the faster bit rate speed having a lower average of power consumed and on top of that a shorter period receiving data transmission. 5.2.1.2 BR/EDR Figure 18: BR/EDR Total mC used When comparing the different bit rate speeds of BR/EDR looking at Figure 18 a trend can also be seen where the least coulomb used tends to be the modulation with the highest speed. Furthermore looking at Table 2, Table 3 and Table 4 a trend can be seen 0 10 20 30 40 50 60 50k 100k 250k 350k 500k mC Bytes sent BLE 1Mb/s BLE 2Mb/s 0 20 40 60 80 100 120 140 160 50k 100k 250k 350k 500k mC Bytes sent Classic 1Mb/s Classic 2Mb/s Classic 3Mb/s 31 where the modulations who support higher speeds will actually draw more average effective power. The speed will therefore be the factor that in this case makes the difference as it does have a shorter period of time that the data has to be sent compared the slower modulations. 5.2.1.3 Compared When comparing the two Bluetooth technologies no breaking point can be found. For the experiments conducted BLE tends to show a much higher average under receiving data transmission. The modulation for BR/EDR with 1 Mb/s is the only BR/EDR that has a higher power consumed, but when comparing its average effective mA used it has the lowest of all the modulations. 5.2.2 Research question 2
• Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on total amount of data sent? 5.2.2.1 BLE The next two figures display the millicoloumb used per byte sent during BLE communication. Figure 19: BLE LE 1M PHY mC used per byte 0,000096 0,000098 0,0001 0,000102 0,000104 0,000106 0,000108 0,00011 50000 100000 250000 350000 500000 Bytes sent mC used per byte 32 Figure 20: BLE LE 2M PHY mC used per byte When comparing the mC used per byte between the LE 1M PHY from Figure 19 and LE 2M PHY from Figure 20 for BLE no real pattern could be found. For BLE LE 1M PHY the collected data seems to be very varied with no real pattern. For BLE LE 2M PHY the collected data shows a higher mC used for 50k bytes sent. When sending more data than 50k bytes a much more stable and lower mC used per byte. BLE LE 1M PHY had an average of 0.10674 μC used per byte and LE 2M PHY had 0.05529 μC. 5.2.2.2 BR/EDR The next three figures display the mC used per byte sent during BR/EDR communication. Figure 21: BR/EDR 1 Mb/s mC used per byte 0,000054 0,0000545 0,000055 0,0000555 0,000056 0,0000565 0,000057 50000 100000 250000 350000 500000 Bytes sent mC used per byte 0,0003005 0,000301 0,0003015 0,000302 0,0003025 0,000303 0,0003035 0,000304 0,0003045 0,000305 0,0003055 50000 100000 250000 350000 500000 Bytes sent mC used per byte 33 Figure 22: BR/EDR 2 Mb/s mC used per byte Figure 23: BR/EDR 3 Mb/s used per byte When looking at the mC used per byte for BR/EDR 1 Mb/s from Figure 21, BR/EDR 2 Mb/s from Figure 22 and BR/EDR 3 Mb/s from Figure 23 a pattern can be seen. The average for 50k bytes will always be one of if not the highest. When bytes sent increase the mC used per byte sent will also start to increase, this goes for all the different modulations. BR/EDR 1 Mb/s had an average of 0.30486 μC used per byte, 2 Mb/s had 0.04364 μC and 3 Mb/s had an average of 0.037249 μC. 5.2.2.3 Compared When comparing the mC used per byte between BR/EDR and BLE, the first noticeable difference is how BLE has no real pattern unlike BR/EDR, which clearly shows that when increasing the total amount of bytes sent, the mC per bytes sent also increases. Whereas the graphs for BLE look more random. BLE has a higher average energy 0,0000415 0,000042 0,0000425 0,000043 0,0000435 0,000044 0,0000445 50000 100000 250000 350000 500000 mC used per byte 0,0000362 0,0000364 0,0000366 0,0000368 0,000037 0,0000372 0,0000374 0,0000376 0,0000378 0,000038 50000 100000 250000 350000 500000 mC used per byte 34 consumption per byte than BR/EDR, which means that BR/EDR is in general more energy efficient than BLE and thus no breaking point can be found. 5.2.3 Research question 3
• How can a circuit be built to accurately measure the required data to validate and verify the two earlier research questions? By utilizing the DSR method and going through multiple stages of an artifact, developing, and improving the artifact in many iterations, the data should become more accurate. Looking at the first artifact, displayed by Figure 2, it was a very simple circuit with an oscilloscope connected to it to measure the electrical current passing through. As shown in Figure 24, the data measured had a lot of noise and may be affected by burden voltage. Thus, the artifact needed improvements. The next artifact, described in Figure 3, was very similar to the first one, using a current probe instead of a regular voltage probe. As seen in Figure 29, the results of this artifact were a lot of added noise to the data and the problem with burden voltage still exists. Figure 4 describes the third iteration of the artifact, where instead of a circuit, the PPK2 is used to power the Bluetooth module and read the electrical current used by it. The PPK2 were chosen because of its ability to read measurements all the way down to 200nA and up to 1A (Nordic semiconductor, u.d.). The software used for the PPK2 is shown in Figure 30, and the data is presented like a simulated oscilloscope. The data is easier to read than previous iterations and the risk of burden voltage is eliminated. In the fourth iteration, the Bluetooth module, Jody-W164, was replaced with the Jody-W263 module, to support the Bluetooth 5.0 technology and gain access to the newest BLE LE 2M PHY. The set-up, as shown in Figure 5, was also different from the previous iteration, as the computer had to power the Bluetooth module and the PPK2 had to be serially connected with the USB-cable to read the data. 35

6. Discussion This chapter will discuss the methods used and the results of the experiments conducted by comparing them to previous studies. 6.1 Result discussion The thesis needed to answer three research questions to fulfil the purpose of the study. 6.1.1 Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on bit rate speed? Looking at the empirical data collected, and the analysis done, it is possible to see the difference in energy consumption between the technologies. Even though the expected bit rate speeds for BR/EDR were not accomplished, the measured results of BR/EDR were still better than the measured results of BLE. The results of BLE shows that BLE is in average much less energy efficient than BR/EDR. The average idle electrical consumption with BLE is 33.67mA compared to 35.14mA with BR/EDR. Even with this difference, BR/EDR is more energy efficient when receiving data. By looking at the graphs, it is possible to see that there is no breaking point between the two technologies when comparing their energy consumption in correlation to the bit rate speed. One reason that could have affected the result is the change of the connection interval between sent packages. To speed up the throughput for BLE a connection interval can be configured to send more packages within a smaller period. This means that the module will almost constantly receive packets and have smaller intervals of idle. And thus, if having a longer connection interval, the average energy consumption will decrease, but the total energy consumed will increase over time. Looking at the Bluetooth 5.0 core specifications the reason why BR/EDR 1 Mb/s received data so slow depended on the packets type sent. The data packet defaulted to DM1 which meant that the max rate would be 108.8 kb/s, even though we had DH5 activated that should give us a max rate of 433.9 kb/s. This means that if the right packet would have been used it could of have had at least, in theory, one quarter of the time used for receiving the same length of packets. This would of have meant an even better energy efficient could of have been measured. The BR/EDR 2 Mb/s had used the 2- DH5 packet with its speed of a max rate of 869.1 kb/s and had a respectable bit rate speed achieved of 786.4 kb/s. The BR/EDR 3 Mb/s had also used the wrong data packet type. BR/EDR 3 Mb/s had used the data packet type of 3-DH3 which meant a max rate of 1177.6 kb/s. If the data packet type would have been the correct one the packet type of 3-DH5 would of have been used. The speed for the 3-DH5 packet would have given us a max rate of 1306 kb/s. As our experiments gave us a speed of 1047.4 kb/s, a speed increase of at least 100kb/s could have been possible which would have given us a slightly better effective energy consumption. This conclusion could be made after looking at the user payload for each of the data packets and then comparing the speeds which matches the core specifications with only a small percent of bit rate speed difference. Comparing the two BLE modulations a faster speed does make it more energy efficient. When looking at the average mA drawn a clear advantage for LE 2M PHY is seen. With 36 LE 1M PHY having on average a 7.8% higher average mA drawn compared to LE 2M PHY and having a slower bit rate speed by 400 kb/s it stands no chance. The LE 2M PHY will therefore always have a shorter period when receiving the packets sent and on top of that draw less on average mA. When looking at the research question with the empirical data collected and the analysis, there was no breaking point that could be found. Thus, because BR/EDR is more energy efficient at sending data with lower mA drawn on average and having higher speeds achieved. When comparing BR/EDR 1 Mb/s to both speeds of BLE it also draws less mA on average but falls short because of its low speed and must receive data for a much longer period. 6.1.2 Where is the breaking point when comparing the energy consumption efficiency between Bluetooth Classic and Bluetooth Low Energy, depending on total amount of data sent? Further investigating the empirical data and analysis, it is possible to see a correlation between the total amount of data sent and the energy consumption through how much every byte consumed. When sending more data with BR/EDR a slow increase could be seen while with BLE it did not show clear signs with LE 1M PHY to make any difference. With BLE LE 2M PHY it showed a more stable but next to none changing mC used per byte. By looking at the charts Figure 19 to Figure 23 there is a clear difference between how much mC BLE uses per byte compared to BR/EDR. BLE uses at an average more coulomb per byte than BR/EDR. BLE LE 1M PHY use 0.10674 μC at an average, and LE 2M PHY use 0.05529 μC, as compared to BR/EDR 1Mb/s, 2Mb/s and 3Mb/s use 0.30486 μC, 0.04364 μC and 0.03725 μC. By looking at this data it is valid to say that no breaking points can be found because BLE were in general less energy efficient than BR/EDR. 6.1.3 How can a circuit be built to accurately measure the required data to validate and verify the two earlier research questions? The requirement for the artifact was to have as little disturbance as possible to measure valid data that would not be faulty in any way. By using the DSR method an artifact that could be used to measure accurate data was created. The development of the artifact was rather difficult as there were a lot of factors to think about to lower the risk of disturbance and get accurate results. After four iterations we had the artifact needed to measure and collect our data. The last iteration was using the Jody-W263 Bluetooth module together with the PPK2 connected to a computer. This artifact had eliminated the use of an actual circuit with measurements taken directly over the resistor, which removes the risk of burden voltage and other risks such as temperature or noise that would affect the data. Another study measuring the data transmission efficiency in BLE (Bulić, Kojek, & Biasizzo, 2019) used the first version of the PPK for measurement of data. As the PPK they used has a resolution of 200 nA we believe this can further increase our measurements validity as the PPK2 has a resolution of 100nA. 37 6.2 Method discussion For this study we decided to use experimentation and observation with help of DSR as our methods of choice to collect and analyse empirical data. We believe that this choice of method was the most relevant one as the purpose of this study was to find the breaking point where the energy consumption of BLE becomes worse than BR/EDR. And to achieve this purpose, many iterations of both hardware and software had to be tested and investigated. The methods we used gave us the relevant data needed to answer the research questions. By using the experimentation and observation method we could gather real life data and observe the data to see any correlations between variables which gave us relevant data. The DSR method then provided us with reliable data that we then could trust for the experiments and after multiple iterations of artifatcts a reliable circuit was made. By conducting experiments with multiple iterations, a higher validity and reliability has been achieved. This is because more iterations with similar results will reduce the risk of corrupt data and instead prove that the data gathered is valid and no external or internal sources have faulted the data. Although we did not achieve the expected bit rate speed when measuring BR/EDR, we still believe that we conducted the experiments correctly since we used the correct modulations for BR/EDR and the results we got would not have been very different if we reached expected throughput speeds. To make sure that the right modulations were used the HCI dumps had to be checked for both BLE and BR/EDR. 38
7. Conclusions and further research This chapter summarizes the study and presents implications and suggested further research. 7.1 Conclusions The purpose of this study was to research and investigate the potential existence of a breaking point where BLE becomes less efficient than BR/EDR in terms of energy consumption. The results from the conducted experiments show that BLE has overall worse energy consumption efficiency than BR/EDR and that there is no breaking point since BLE is shown to be worse. 7.1.1 Implications The results from this study show the differences between BLE and BR/EDR and could help the industry, more specifically car manufacturers, decide whether to use BR/EDR or BLE as their Bluetooth technology in their cars. 7.2 Further research Due to time limitations and other limitations mentioned in the chapter Scope and limitations, further research exists. Because we used a licensed Bluetooth stack that is not open for public use, similar experiments but with an open-source stack would be interesting to analyse and verify that there are no differences between a licensed Bluetooth stack and an open-source stack. The Jody-W263 we used for our experiments is a Bluetooth module focused on automotive use and thus may have different results than when using a more common Bluetooth module, such as a phone which is more heavily focused on small power consumption. Thus, a relevant extension to this study would be to test multiple different Bluetooth modules to see if there may be any differences. 39
8. References Bergelin, J., & Ericsson, F. (2019). Dataöverföring med Bluetooth 5. Jönköping: diva- portal. Bluetooth. (2012, July 24). Serial Port Porfile 1.2. Retrieved from Bluetooth: https://www.bluetooth.com/specifications/specs/serial-port-profile Bluetooth. (2022, March 16). Bluetooth. Retrieved from Bluetooth® technology media information: bluetooth.com/media/ Bluetooth. (2022, 04 19). Bluetooth Technology Overview. Retrieved from Bluetooth: https://www.bluetooth.com/learn-about-bluetooth/tech-overview/ Bluetooth. (2022, April 19). Topology Options. Retrieved from Bluetooth: https://www.bluetooth.com/learn-about-bluetooth/topology-options/ Bluetooth SIG, Inc. (2016, 12 06). bluetooth. Retrieved from Core Specification 5.0 - Bluetooth® Technology Website: https://www.bluetooth.com/specifications/specs/core-specification-5/ Bulić, P., Kojek, G., & Biasizzo, A. (2019). Data Transmission Efficiency in Bluetooth Low Energy Versions. Ljubljana: Sensors. doi:https://doi.org/10.3390/s19173746 CD & Power. (n.d.). gotpower. Retrieved from Brownouts vs. Blackouts. What's the difference?: https://www.gotpower.com/brownouts-and-blackouts/ Daire, A. (2005). Low-Voltage Measurement Techniques. Cleveland: Keithley Instruments, Inc. Direct Energy. (n.d.). directenergy. Retrieved from What's the Difference Between a Blackout and a Brownout? | Direct Energy: https://www.directenergy.com/learning-center/difference-between-blackout- brownout Donovan, J. (den 01 December 2011). Bluetooth Goes Ultra-Low-Power. Hämtat från Digikey: https://www.digikey.com/en/articles/bluetooth-goes-ultra-low-power EEPower. (n.d.). EEPower. Retrieved from Shunt Resistor | Resistor Applications | Resistor Guides: https://eepower.com/resistor-guide/resistor- applications/shunt-resistor/# Knowit. (2022, March 21). About us. Retrieved from Knowit: https://www.knowit.eu/about-knowit/ Knowit. (2022, March 20). blueGO. Retrieved from Knowit: https://www.knowit.eu/services/connectivity/product-service- development/bluego/ Maker.io Staff. (2022, March 20). Digikey. Retrieved from Understanding the Basics of Infrared Communications: https://www.digikey.com/en/maker/blogs/2021/understanding-the-basics-of- infrared-communications 40 Marcel, J. (2018, June 20). How the Bluetooth Community Revolutionized Data Transfer. Retrieved from Bluetooth: https://www.bluetooth.com/blog/how- bluetooth-revolutionized-data-transfer/ NI. (2018). Burden Voltage. Retrieved from NI: https://zone.ni.com/reference/en- XX/help/370384V-01/dmm/burden_voltage/ Nordic semiconductor. (u.d.). Power Profiler Kit II . Hämtat från Nordicsemi: https://www.nordicsemi.com/Products/Development-hardware/Power- Profiler-Kit-2 Ren, K. (2017, February 20). Higher Speed How Fast Can It Be? Retrieved from Bluetooth: https://www.bluetooth.com/blog/exploring-bluetooth-5-how-fast- can-it-be/ Siekkinen, M., Hiienkari, M., K. Nurminen, J., & Nieminen, J. (2012). How low energy is bluetooth low energy? Comparative measurements with ZigBee/802.15.4. 2012 IEEE Wireless Communications and Networking Conference Workshops (WCNCW), 232-237. doi:10.1109/WCNCW.2012.6215496 Tosi, J., Taffoni, F., Santacatterina, M., Sannino, R., & Formica, D. (2017). Performance Evaluation of Bluetooth Low Energy: A Systematic Review. Basel, Switzerland: Sensors. doi:https://doi.org/10.3390/s17122898 vom Brocke, J., Hevner, A., & Maedche, A. (2020). Introduction to Design Science Research. SpringerNature. 41
9. Appendix Figure 24: Probe 50k bytes sent. Figure 25: Probe 500k bytes sent. 42 Figure 26: Probe 1M bytes sent. Figure 27: Current probe 50k bytes sent. 43 Figure 28: Current probe 500k bytes sent. Figure 29: Current probe 1M bytes sent. 44 Figure 30: PPK2 50k bytes sent. Figure 31: PPK2 500k bytes sent. Figure 32: PPK2 1M bytes sent.

---

https://academy.nordicsemi.com/courses/bluetooth-low-energy-fundamentals/

Skip to content

Register Log in

Bluetooth Low Energy Bluetooth Low Energy Fundamentals 9.1/10 (2390) 7904already enrolled Course description

The Bluetooth Low Energy Fundamentals course is an online, self-paced course that focuses on teaching the basics of Bluetooth LE using Nordic Semiconductor devices (nRF54, nRF53, and nRF52 Series). Through hands-on learning, participants will learn how to create a Bluetooth LE prototype and establish a wireless, unidirectional and bidirectional data channel between two Bluetooth LE devices in a short period of time.

Upon completion of the course, participants will have a thorough understanding of the Bluetooth LE protocol and its layers, as well as knowledge of available APIs in the nRF Connect SDK, which is based on the Zephyr RTOS. Additionally, participants will have gained practical experience in configuring Bluetooth LE advertisements and connections, as well as insight into securing Bluetooth LE connections and inspecting packets over the air using nRF Sniffer

What you'll learn

Develop a solid comprehension of the most up-to-date Bluetooth LE application architecture. Acquire knowledge of Bluetooth LE advertising and its functionalities. Learn about Bluetooth LE connections and how to exchange data over Bluetooth LE. Gain insights on how to secure Bluetooth LE connections. Familiarize yourself with the Bluetooth LE APIs in the nRF Connect SDK. Learn about tools (nRF Connect for Mobile, nRF Sniffer) for debugging Bluetooth LE applications. Apply your learning through hands-on exercises to establish unidirectional and bidirectional data channel between two Bluetooth LE-enabled devices.

Who should enroll in this course?

You should enroll to this course if you are interested in learning about the Bluetooth LE protocol and want to build a Bluetooth LE prototype using Nordic Semiconductor’s products. The course is aimed to embedded software engineers, firmware developers, or anyone interested in microcontrollers and embedded systems. Estimated time An estimate of about eight to ten hours in total is needed to finish this course. The actual time needed to finish the course will highly depend on your technical background and experience. Learning Objectives Examine the Bluetooth LE protocol and understand the role of its different layers Master Bluetooth LE advertising Learn about Bluetooth LE connections Examine the available options for data exchange in Bluetooth LE Gain insights on how to secure Bluetooth LE connections Examine Bluetooth LE packets over the air Start course Details Fundamental level 6 lessons 8-10 hours to complete Certificate of completion Prerequisites

Basic knowledge in the C programming language. Some experience in developing software for embedded systems. nRF Connect SDK Fundamentals course (Required) Lesson 1 (Recommended) Lesson 2 – Lesson 8

Hardware Requirements

Any of the following development kits: nRF54LM20 DK, nRF54L15 DK, nRF5340 DK, nRF52840 DK, nRF52833 DK, nRF52 DK A smartphone or a tablet (Android 4.3 or later and iOS version 16.0 or later) Special hardware requirements for Lesson 6

System Requirements

A computer running Windows, macOS, or Linux An up-to-date web browser

Software Requirements​

nRF Connect for Desktop nRF Command Line Tools Visual Studio Code nRF Connect for Mobile

Supported SDK versions

nRF Connect SDK v3.1.1 – v2.3.0 nRF Connect SDK v3.1.1 or higher is needed for the nRF54LM20 DK nRF Connect SDK v2.8.0 or higher is needed for the nRF54L15 DK

Course Content Expand All Lesson 1 – Bluetooth LE Introduction 5 Topics | 1 Quiz What is Bluetooth LE? GAP: Device roles and topologies ATT & GATT: Data representation and exchange PHY: Radio modes Exercise 1 Lesson 1 quiz Lesson 2 – Bluetooth LE Advertising 7 Topics | 1 Quiz Advertising process Advertising types Bluetooth address Advertisement packet Exercise 1 Exercise 2 Exercise 3 Lesson 2 quiz Lesson 3 – Bluetooth LE Connections 4 Topics | 1 Quiz Connection process Connection parameters Exercise 1 Exercise 2 Lesson 3 quiz Lesson 4 – Data exchange in Bluetooth LE 6 Topics | 1 Quiz GATT operations Services and characteristics Attribute table Exercise 1 Exercise 2 Exercise 3 Lesson 4 quiz Lesson 5 – Security in Bluetooth LE communication 5 Topics | 1 Quiz Pairing process Legacy pairing vs LE Secure Connections Security modes Exercise 1 Exercise 2 Lesson 5 quiz Lesson 6 – Bluetooth LE sniffer 5 Topics | 1 Quiz Sniffing Bluetooth LE packets Setting up nRF Sniffer for Bluetooth LE Exercise 1 Exercise 2 Exercise 3 Lesson 6 quiz Get your Certificate! Nordicsemi.com Nordicsemi.cn Nordicsemi.jp TechDocs DevAcademy DevZone TechWebinars Contact Us Privacy Policy Terms of Service Sitemap Copyright © 2025 Nordic Semiconductor. All rights reserved