Scheduling with Outdated CSI: Effective Service Capacities of Optimistic vs. Pessimistic Policies
2012 (English)In: 2012 IEEE 20th International Workshop on Quality of Service (IWQoS), IEEE conference proceedings, 2012, 1-9 p.Conference paper (Refereed)
The concept of the effective service capacity is an analytical framework for evaluating QoS-constrained queuing performance of communication systems. Recently, it has been applied to the analysis of different wireless systems like point-to-point systems or multi-user systems. In contrast to previous work, we consider in this work slot-based systems where a scheduler determines a packet size to be transmitted at the beginning of the slot. For this, the scheduler can utilize outdated channel state information. Based on a threshold error model, we derive the effective service capacity for different scheduling strategies that the scheduler might apply. We show that even slightly outdated channel state information leads to a significant loss in capacity in comparison to an ideal system with perfect channel state information available at the transmitter. This loss depends on the risk-level the scheduler is willing to take which is represented by an SNR margin. We show that for any QoS target and average link state there exists an optimal SNR margin improving the maximum sustainable rate. Typically, this SNR margin is around 3 dB but is sensible to the QoS target and average link quality. Finally, we can also show that adapting to the instantaneous channel state only pays off if the correlation between the channel estimate and the channel state is relatively high (with a coefficient above 0.9).
Place, publisher, year, edition, pages
IEEE conference proceedings, 2012. 1-9 p.
Communication Systems Telecommunications
IdentifiersURN: urn:nbn:se:kth:diva-136840DOI: 10.1109/IWQoS.2012.6245966ScopusID: 2-s2.0-84866594042ISBN: 978-146731298-1OAI: oai:DiVA.org:kth-136840DiVA: diva2:677315
2012 IEEE 20th International Workshop on Quality of Service, IWQoS 2012; Coimbra, Portugal, 4-5 June 2012
QC 201312182013-12-092013-12-092014-01-07Bibliographically approved