Admission control of delay bounded traffic in cellular networks

In this thesis we consider problems arising in delivering delay bounded traffic to mobile users of cellular networks. Such traffic is important to support (for example, for delivering multimedia streaming services), yet challenging to manage, since it typically requires relatively high data rates that consume significant wireless system resources. Moreover, users expect uninterrupted operation while roaming within the coverage area. The problems formalized in the thesis are investigated in the context of managing two types of cellular networks: (1) networks that support multi user access through time division multiplexing where a fixed number of channels is allocated to serve the streaming requests, and (2) networks that support multi user access through code division multiple access (CDMA). Such networks are characterized by a soft capacity aspect. For networks of the first type, we devise admission control (CAC) mechanisms that keep track of network state at any instant by utilizing scheduling mechanisms that take into account delay constraints on individual traffic connection requests. Two types of scheduling mechanisms are considered in the thesis for the above purpose: non-preemptive scheduling that assumes that a connection request is served to completion without service interruption, and preemptive scheduling that aims at achieving higher throughput by allowing service preemption. In both cases, the thesis develops frameworks for the devised CAC and the underlying scheduling mechanism, present quantitative analysis of the designed schedulers, and evaluate the performance of the devised frameworks by simulation. A novel contribution of the thesis is the design and analysis of CAC architectures for the first type of networks to serve delay bounded traffic. For networks of the second type, we devise a CAC mechanism that keeps track of network state at any instant by keeping track of both intra-cell and inter-cell mobility of served users in order to estimate the cell overload probability after a prediction interval in the future. Such CAC architecture has been devised in the literature for rate sensitive (but not particularly delay bounded) traffic. A novel aspect of the thesis is on extending the architecture to our present context of serving delay bounded traffic in soft capacity networks.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.