Opportunistic Multicast Scheduling With Resource Fairness Constraints In Cellular Networks

Multicast is emerging as an important means of transmitting the same content to multiple receivers so as to efficiently utilize the overall network resources. Being motivated by recent advances in erasure codes and fountain codes, opportunistic multicast scheduling algorithms have been proposed recently to better balance the tradeoff between multi-user diversity and the multicast gain. The data-rate that each user can support varies during different time slots, the opportunistic multicast scheduling algorithms select a different multicast group for transmission in each time slot at a rate supported by the worst user in the selected group. While opportunistic multicast scheduling imposes some new challenges in terms of fairness, not only among different multicast groups, but also among users that belong to the same multicast group.Designing multi-user scheduling algorithms is a very challenging problem in wireless networks. Due to the time-varying and user-dependent nature of channel conditions, scheduling algorithms need to dynamically allocate resources based on both the channel states and the user demands to achieve higher utilization of wireless resources. Opportunistic scheduling has recently emerged as a promising technique for providing highrate data services in cellular data networks. Opportunistic scheduling can effectively increase network throughput by exploiting multi-user diversity inherent in wireless networks. However, the potential to exploit higher data throughput in an opportunistic way also introduces an important issue between resource efficiency and fairness tradeoff among different users.In this paper, we propose an opportunistic multicast scheduling algorithm that maximizes the overall network throughput by exploiting time-varying channel conditions while taking into account the resource fairness constraints among all groups. The optimum transmission rate for each group is selected based on the current channel data rate and average received throughput of each user to achieve the best tradeoff between throughput and fairness. We then prove the optimality of the multicast scheduling algorithm and show that our algorithm does not deteriorate any group's performance to improve the overall network throughput compared with the non-opportunistic scheduling algorithm. Through extensive simulations we demonstrate that the proposed opportunistic multicast scheduling algorithm improves network resource efficiency under fairness constraints.