Real-time remote processing enabled by high speed Ethernet
(2023) EITM02 20231Department of Electrical and Information Technology
- Abstract
- A growing trend within the technologies acting as enablers for 6g, such as Massive MIMO and Large Intelligence Surfaces, is benefiting from both the communication and the positioning aspects that they can provide. As these kinds of systems are employing a large number of arrays which provide high amounts of data, a distributed hardware approach having near-antenna processing is explored in this work. The emulated set-up consists of two FPGA boards, where one is mimicking
the local processing which would take place near the LIS panels, while the other aggregates all the data gathered from the panels to have joint processing. The channel state information is gathered through four LIS panels and is processed locally with the help of deep... (More) - A growing trend within the technologies acting as enablers for 6g, such as Massive MIMO and Large Intelligence Surfaces, is benefiting from both the communication and the positioning aspects that they can provide. As these kinds of systems are employing a large number of arrays which provide high amounts of data, a distributed hardware approach having near-antenna processing is explored in this work. The emulated set-up consists of two FPGA boards, where one is mimicking
the local processing which would take place near the LIS panels, while the other aggregates all the data gathered from the panels to have joint processing. The channel state information is gathered through four LIS panels and is processed locally with the help of deep neural networks in order to achieve the position of a
user. The next step is sending this data onwards to the board which acts as the central processing unit in order to fuse it, thus achieving a better estimation of
the position.
This thesis is proposing a hardware implementation focusing on different optimization techniques that could be used in order to achieve a low-latency and high-throughput system. For real-time applications where latency is critical, such as, for example, autonomous vehicles – a good approach are hardware accelerators tailored exclusively to the function which needs to be implemented. Finally, the communication between hardware units is also managed, with the help of an
Ethernet link. (Less) - Popular Abstract
- With the rise of Internet of Things, which enables communications between smart devices, people have been yearning and thinking about the possibilities that may arise with controlling IoT in real time. This can only be possible by achieving more speed and accuracy within the human-machine interaction. In the technical world, this means combining different growing technologies such as Large Intelligent Surfaces, Deep Neural Networks or Systems on Chips.
As the developing wireless technology is approaching 6G, the need for hardware supporting this emerging technology is directly increasing. 6G is promising high communication data-rates and high positioning accuracy, which needs to be supported in some manner by the real-time systems. This... (More) - With the rise of Internet of Things, which enables communications between smart devices, people have been yearning and thinking about the possibilities that may arise with controlling IoT in real time. This can only be possible by achieving more speed and accuracy within the human-machine interaction. In the technical world, this means combining different growing technologies such as Large Intelligent Surfaces, Deep Neural Networks or Systems on Chips.
As the developing wireless technology is approaching 6G, the need for hardware supporting this emerging technology is directly increasing. 6G is promising high communication data-rates and high positioning accuracy, which needs to be supported in some manner by the real-time systems. This work aims to emulate a software distributed system proposed in "Positioning for distributed large intelligent surfaces using neural network with probabilistic layer", in hardware. The system consists of four panels situated on each wall of a room which capture the information of how the signal propagates from the user to the panels. This is called CSI -- Channel state Information. These signals have to be processed rapidly and accurately to get a precise location before, for example, the user moves. In hardware, one has to consider the limitations of a computationally-intensive algorithm which are often about how much data can be stored at once, how many operations can be performed in parallel or with which accuracy the numbers need to be represented. Moreover, the problem of transmitting data from one panel is also tackled. As the architecture of a distributed system implies having different hardware units which are processing data in parallel which is then sent to a central processing unit, the communication between them will also represent a challenge. All of these dilemmas are handled within this work.
The Field Programmable Gate Arrays - FPGAs are an excellent platform for developing new architectures, as it allows for fast prototyping. As the provided FPGAs boards include an Ethernet port, this protocol has been chosen to transmit the data. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9126479
- author
- Iancu, Dumitra LU and Tinnerberg, Lina
- supervisor
- organization
- course
- EITM02 20231
- year
- 2023
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Large Intelligent Surfaces, Deep Neural Networks, FPGA, VHDL, Ethernet
- report number
- LU/LTH-EIT 2023-931
- language
- English
- id
- 9126479
- date added to LUP
- 2023-06-21 13:25:17
- date last changed
- 2023-06-21 13:25:17
@misc{9126479, abstract = {{A growing trend within the technologies acting as enablers for 6g, such as Massive MIMO and Large Intelligence Surfaces, is benefiting from both the communication and the positioning aspects that they can provide. As these kinds of systems are employing a large number of arrays which provide high amounts of data, a distributed hardware approach having near-antenna processing is explored in this work. The emulated set-up consists of two FPGA boards, where one is mimicking the local processing which would take place near the LIS panels, while the other aggregates all the data gathered from the panels to have joint processing. The channel state information is gathered through four LIS panels and is processed locally with the help of deep neural networks in order to achieve the position of a user. The next step is sending this data onwards to the board which acts as the central processing unit in order to fuse it, thus achieving a better estimation of the position. This thesis is proposing a hardware implementation focusing on different optimization techniques that could be used in order to achieve a low-latency and high-throughput system. For real-time applications where latency is critical, such as, for example, autonomous vehicles – a good approach are hardware accelerators tailored exclusively to the function which needs to be implemented. Finally, the communication between hardware units is also managed, with the help of an Ethernet link.}}, author = {{Iancu, Dumitra and Tinnerberg, Lina}}, language = {{eng}}, note = {{Student Paper}}, title = {{Real-time remote processing enabled by high speed Ethernet}}, year = {{2023}}, }