12月26号在新主楼第四会议室举行了实验室Workshop，这次Workshop有5位老师和同学进行讲解，内容如下：
EE-Delay Relationship in Wireless Systems with Statistical QoS Requirement
When transmitting random traffic over wireless channels, it is well-known that a tradeoff exists between transmit power and average delay. In this paper, we consider statistical QoS requirement characterised by delay bound and delay bound violation probability, and study the relationship between the energy efficiency (EE) and the delay bound. We show that if the total power consumption including both transmit and circuit powers of a base station (BS) linearly increases with the required average service rate, the EE-delay tradeoff will vanish. By taking massive multi-input-multi-output (MIMO) system as an example, we show that the total power linearly increases with the average rate when transmit power and bandwidth are jointly allocated. The EE-delay relationship and the optimal queue state dependent policy are derived, and the impacts of BS idling and resource constraints are then addressed. A closed-form optimal policy is provided for compound Poisson source served by massive MIMO system. The results show that the EE-delay curve includes a tradeoff region and a non-tradeoff region, and the non-tradeoff region increases with the number of transmit antennas.
Low Complexity Precoding for Massive MIMO
In the massive MIMO downlink transmission scenarios, ZFBF and MRT precoders are widely applied. For ZFBF precoder, the increasing number of transmitting antennas at base station leads to high computational complexity in the procedure of matrix multiplication; For MRT precoder, it is of low computational complexity but its performance may be affected by angle spread. Based on the above, we consider the low-complexity precoder design on angle domain: we transform the array-domain channel into angle domain and reconstruct the channel projection matrix with sparsity to lower computaional complexity and guarantee performance.
