CMCVT: A Concurrent Multi-Channel Virtual Transceiver

• Published in: International journal of electronics and communications Vol. 120; p. 153230
• Main Authors: Aslam, Muhammad, Jiao, Xianjun, Liu, Wei, Moerman, Ingrid
• Format: Journal Article
• Language: English
• Published: Elsevier GmbH 01-06-2020
• Subjects: FPGA; IoT; Multi-channel transceiver; Radio virtualization; SDR; SDR; Multi-channel transceiver; FPGA; Radio virtualization; IoT
• ISSN: 1434-8411; 1618-0399
• DOI: 10.1016/j.aeue.2020.153230

State-of-the-art wireless Gateways (GW) used in Internet of Things (IoT) offer a single channel radio link, which limits the capabilities of the IoT network controlled by the GW, as the GW can only use a single channel at a time to communicate with the end-device(s). The quality of service (e.g., aggregate throughput, latency) offered by a single channel GW could be substantially improved by employing a multi-channel transceiver, which is capable of transmitting/receiving data on different radio channels simultaneously, particularly for larger wireless networks. However, current solutions available in both research and commercial communities only offer multi-channel receiver capabilities, and do not incorporate the multi-channel transmitter part. In addition, in terms of implementation, these multi-channel receivers duplicate single-channel hardware functionality. In this paper, for the first time, a novel concurrent multi-channel virtual transceiver is introduced. The virtual transceiver offers multi-channel capabilities and uses the same single-hardware hardware implementation for the Physical (PHY) layer by employing the virtualization technique. This new virtual transceiver concept is demonstrated for an IEEE 802.15.4 based 8 × 8 channel transceiver, implemented on an Field Programmable Gate Array (FPGA) of a modern Software Defined Radio and is compared with the existing duplication approach. The duplication approach consumes 9008 LUTs, and 12120 FFs, whereas the proposed approach occupies only 2959 LUTs and 2105 FFs, saving 67.15% LUTs and 82.63% FFs in comparison with the duplication approach. The experimental results reveal that the virtual transceiver provides the same performance (e.g., receiver sensitivity of −98.5dBm) as the transceiver achieved by duplicating the PHY layers but consumes much less hardware resources.