Diversity Enabled Low-Latency Wireless Communications With Hard Delay Constraints

The emerging next generation Ultra-Reliable and Low-Latency Communications (xURLLC) is expected to play a central role in supporting mission-critical mobile applications because it holds the promise of improving the Quality-of-Service (QoS) substantially. However, it is quite challenging to satisfy the hard delay constraint in harsh wireless environments due to sporadic deep fades, especially when the average power is strictly limited. In this paper, we aim at assuring hard delay constraints with the aid of frequency or spatial diversity techniques. To this end, we focus on both parallel and multiple-input-multiple-output (MIMO) fading channels, in which time domain power adaptation is exploited to provide just-in-time services (JITS). It is shown that the hard delay constraint can be satisfied with a finite average power when the frequency or spatial diversity gains are no less than two. By adopting the implicit function theorem, we reveal the relationship between the required average power, the delay constrained throughput, and the outage probability without power adaptation. Furthermore, by adopting Ferrari's solution to fourth order algebraic equations, we show that hard delay constrained transmission is feasible even when the sub-channels in the frequency and spatial domains are highly but not fully correlated.