Abstract
Owing to limited resources, it is hard to guarantee minimum service levels to all users in conventional cellular systems. Although global cooperation of access points (APs) is considered promising, practical means of enhancing efficiency of cellular systems is by considering distributed or clustered jointly processed APs. The authors present a novel `quality of service (QoS) balancing scheme' to maximise sum rate as well as achieve cell-based fairness for clustered jointly processed cellular multiple access channel (referred to as CC-CMAC). Closed-form cell level QoS balancing function is derived. Maximisation of this function is proved as an NP hard problem. Hence, using power-frequency granularity, a modified genetic algorithm (GA) is proposed. For inter site distance (ISD) <; 500 m, results show that with no fairness considered, the upper bound of the capacity region is achievable. Applying hard fairness restraints on users transmitting in moderately dense AP system, 20% reduction in sum rate contribution increases fairness by upto 10%. The flexible QoS can be applied on a GA-based centralised dynamic frequency planner architecture.