Resource allocation in moving small cell network

Abstract

The exponentially increasing capacity demand and ubiquitous coverage requirement of cellular users has shifted the cellular operators towards heterogeneity in the network. Heterogeneity promises to meet the capacity demand and coverage requirement by creating small cells on top of the existing macrocells. Small cells are coverage areas served by low powered base stations, called small cell access points (SAPs). SAPs can be deployed to provide good Quality-of-Service (QoS) to users in areas where macrocell antenna’s signal quality is low. Users in moving vehicles such as trains and buses are among those who have been suffering from low QoS. This paves the way for SAPs to be deployed in vehicles and we call them as Moving-SAPs (M-SAPs).

Since SAPs are added on top of the existing macrocell, the co-channel interference makes the resource allocation a challenging problem. Furthermore, with the addition of M-SAPs, the dynamic interference relationship exacerbates the problem. Intelligent resource allocation schemes are required to fully utilize the benefits of deploying M-SAPs in the network. In this thesis, we studied the resource allocation problem in an uplink orthogonal frequency division multiple access (OFDMA) cellular network with M-SAPs deployed. We considered OFDMA due to its inherent robustness against frequency selective fading. The resources distributed in the network are OFDMA resource blocks (RB) and their assigned transmission powers.

We first studied the allocation of RBs to moving small cell network with deterministic mobility (e.g. M-SAPs deployed in trains). The interference relationship is determined using the mobility model of the SAPs. We proposed to represent the interference relationship in the network as a time-interval-dependent-interference (TIDI) graph. Using the TIDI graph, a cluster-based resource allocation (CBRA) is proposed to allocate RBs to small cells such that the RBs are efficiently utilized in the network.

We then investigated the resource allocation of RBs to a moving small cell network with non-random mobility (e.g. M-SAPs deployed in city buses). Exploiting the headway characteristics of the city buses, we captured the interference relationship between SAPS. We proposed a probabilistic graph-based resource allocation (PGRA) scheme to distribute RBs in the moving small cell networks to efficiently utilize the RBs.

Next, we solved the problem of jointly allocating RBs and transmission power in a small cell network. Owing to the complexity of the problem, we decomposed the problem into the sub-problems to make it more tractable. The sub-problems are a) RB allocation for a fixed power allocation, and b) the power allocation for a fixed RB allocation. The sub-problems are solved separately. Using the results obtained from the sub-problems, we proposed an iterative resource allocation algorithm (IRAA) to compute the RBs and transmission powers assigned to the users in the moving small cell network.

Finally, we studied the backhaul resource allocation in the downlink for the newly arrived SAPs in the network such that the service requirement of the macrocell users and the backhaul of the existing SAPs are satisfied.