Adaptive resource allocation for D-TDD systems in wireless multimedia networks

by 1973- Yun, Jungnam

iii In accordance with the demand for robust and spectrally efficient communications systems design for multimedia communication with heterogeneous traffic in fading channels, various adaptive approaches have been developed, recently. This dissertation examines how adaptive technologies can be used in real systems in order to bring reliable and spectrally efficient connections between wireless communication links. It addresses two different types of adaptation: (1) channel adaptation with adaptive modulation /coding, adaptive antennas, and automatic repeat request (ARQ) schemes; (2) bandwidth adaptation to the dynamic unbalanced network traffic with dynamic time division duplex (D-TDD) scheme. In order to make the vulnerable wireless channel reliable and spectrally efficient, we adopt adaptive modulation/coding and ARQ schemes. We introduce an analytical method, which uses a finite-state Markov chain (FSMC) as an error model, for estimating the performance of adaptive modulation systems (AMS) combined with ARQ schemes in correlated slow fading channels. For the purpose of the throughput performance evaluation of wireless packet networks, fading channels have been assumed to have independent and identically distributed (i.i.d) statistics. This assumption of channel model may be suitable to represent fast fading channels. However, in slow fading channels, error rates of consecutive packets are highly correlated, and we cannot simply assume an independent error process for the purpose of performance evaluation. We propose a multi-state Markov error model for AMS in correlated fading channels, which is also described by a Finite-State Markov Chain (FSMC), and we present throughput estimation methods for AMS combined with ARQ, using the proposed Markov error model. For multimedia services, network traffic becomes heterogeneous; the network traffic is homogeneous for voice traffic only. It has been known that dynamic-TDD (D- TDD) scheme utilizes the spectrum more efficiently, adjusting the ratio of uplink and iv downlink available timeslots adaptively according to the traffic pattern. However, having different switching points in each cell causes severe co-channel interference between base stations. We present a statistical assessment of co-channel interference of D-TDD scheme when switched beam sector antennas are deployed at base stations, and we also present a cooperative co-channel interference avoidance algorithm in TDMA/D-TDD wireless cellular systems. Finally, with these channel and bandwidth adaptations, we propose adaptive resource allocation schemes considering joint parameters in the physical (PHY) layer and the medium access control (MAC) layer. We consider cases without co-channel interference and with co-channel interference, denoting as single and multi- cluster cases. For the single cluster case, we first allocate bandwidths for uplink and downlink based on the requested bandwidths for both links. Secondly, we allocate modulation format to each call based on the estimated channel state; this procedure makes some more channels available in the frame for the calls waiting in the queue. For the multi-cluster case, in addition to resource allocation procedures for the single cluster case, we have to allocate time slots to each call appropriately in a way of avoiding severe co-channel interference. By simulation results, we show that more calls (or users) can be serviced in D-TDD systems than fixed-TDD (FTDD) systems with (or without) adaptive modulation.