Dynamic programming based mobile relay selection algorithm in millimeter-wave broadband 5G network

Abstract

Due to growing wireless traffic demand and bandwidth shortage under 3GHz, radio frequency spectrum within 3-300 GHz, millimeter-wave (mmWave) bands are being spotlighted for the next generation of mobile broadband network. mmWave mobile broadband network is expected to achieve high capacity of several Gbps by utilizing immense bandwidth from hundreds MHz to GHz and beamforming antenna technology based on its property of strong straightness. However, the straightness of mmWave is advantage and also disadvantage at the same time. mmWave is harshly attenuated by dense buildings in metropolitan area. As the result, it is predicted to generate outage zone where the signal does not reach frequently in mmWave network. Nowadays, there are many field tests using mmWave equipment in worldwide. According to 28GHz propagation measurement in Manhattan, New York [1], 16% of tested links are outage when distance between transmitter and receiver is within 20-200m and 57% are outage for 200-500m. As many researchers predicted, severe outage problem is observed and it can greatly inhibit the quality of service (QoS). We should overcome this problem to implement and operate mmWave network as the next generation of mobile network. In the other side, energy efficiency has attracted a lot of attention in network area. With the tremendous growth of wireless data traffic, energy consumption of wireless network will be almost 50% of entire ICT technology [2]. Especially, it is more important issue for mmWave network because of its potential Gbps capacity. Energy efficiency of mmWave network should be much higher than energy efficiency of conventional cellular networks to prevent the terrible amount of energy consumption caused by mmWave network. As we described, in mmWave network, there are two important objectives, reducing outage probability and improving energy efficiency. Relay communication has been studied to solve these problems. In general, relay is network equipment which receives signal from base station (BS) and forwards it to mobile devices. Relay can provide bypass for users in outage zone. Also, it can provide energy-efficient bypass which can substitute wireless link with bad channel quality. Therefore, relaying is fascinate solution for mmWave network considering outage probability and energy efficiency together. In this thesis, we assume that users have mobile devices which can relay to each other. We are trying to reduce outage probability and improve energy efficiency by utilizing mobile relays. If mobile relays greatly improve the performance of mmWave network, it will be the cheapest solution for mmWave network instead of constructing and operating fixed relay stations. Thus, we have combinatorial optimization problem of mobile relay selection: how could we decide the optimal mobile relay selection, who will relay and who will receive it. To solve this problem, at first, we propose novel proportional fairness of energy metric considering outage probability and energy efficiency together. Consequently, we propose dynamic programming based mobile relay selection algorithm. It considers numerous cases of relay selections in efficient way and finds the mobile relay selection which maximizes the proposed metric. As the result, compared to mmWave network without mobile relays, the proposed algorithm improves energy efficiency by 25.6% while reducing outage probability by 28.6%. In addition, it achieves the balanced performance compared to conventional relay strategies which optimizes the one of outage probability or energy efficiency.