Achievable Rate Analysis of Millimeter Wave Channels with Random Coding Error Exponent

With emerging applications, e.g., factory automation, autonomous driving and augmented/virtual reality, there have been increasing technical challenges regarding reliability, latency data rates for existing communication systems. Owing to abundant available bandwidth, millimeter Wave (mmWave) communications can potentially provide reliable with the order of magnitude capacity improvement relative microwave, sub 6 GHz communications. Though are many research results improving throughputs, reliability performance mmWave is still not quite clear, especially finite blocklength regimes. In this paper, we investigate achievable channels by random coding error exponent. Under assumption perfect imperfect channel state information at receiver (CSIR), exact approximate analytical expressions derived capture relationship rate, reliability. Furthermore, show that rate always increases as bandwidth CSIR. However, exists a critical maximizes non-line-of-sight signals CSIR, beyond which will decrease bandwidth. For training symbol length power allocation factor maximizing phase investigated closed-form special cases derived.