Providing High Reliability on Blockage-Prone Higher Frequency Bands With Packet-Level Coding

Higher frequency bands (HFB) such as mmWave and sub-THz are important constituents of next generation wireless communication technology such as 6 G, as they offer large bandwidth and capacity to support high data rates. However, due to the narrow beam width and high penetration loss, such HFB links are easily susceptible to blockage effects caused by moving objects such as humans and cars. Blockage durations are relatively long (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim 500$</tex-math></inline-formula> milliseconds) while some 6G use cases require a high reliability within a much shorter delay, which makes it challenging to support them on HFB. In this paper we use linear packet-level coding, which is a spectrally-efficient low-latency proactive reliability enhancement technique above the physical (PHY) layer, to efficiently utilize the aggregated bandwidth of multiple HFB links to provide the required quality of service (QoS) of high-reliability and delay-critical traffic. To effectively combat the deleterious effects of blockage on the HFB links, we analytically characterize the reliability performance and its limit of packet-level coding in the blockage-prone multi-link environment, and based on which we design optimization algorithms to minimize the number of HFB links and the radio resource needed to achieve the required QoS. Numerical results show that our coding solution requires far less HFB links than the existing packet duplication technique defined in 5G standards, and with low complexity our optimization algorithm can significantly reduce the amount of radio resource needed. Our solution provides a feasible way of harvesting the vast bandwidth of blockage-prone HFB to provide high-reliability low-latency data services, which cannot be practically enabled by traditional techniques such as packet duplication or retransmission.