ABSTRACT: The performance evaluation of the IEEE 802.11 MAC protocol has recently attracted attention from research mainly by means of simulation. Most proposed analytic models focus on the analysis of throughput and MAC transmission delay. However, so far there is no analytic model for the MAC queueing delay, which itself is a significant part of the end-to-end delay. Moreover, all the existing analytical and simulation models overlook an important phenomenon in any practical network - blocking due to finite buffering capacity. To eliminate the packet blocking simply by increasing the capacity of the station queue is not acceptable, because the additional queue delay introduced will cause the failure of multimedia application with strict time constraints. In this paper, for the first time we establish a finite queueing model for the IEEE 802.11 MAC protocol and evaluate the packet blocking probability and the MAC queueing delay.
Vehicular Technology Conference, 2004. VTC 2004-Spring. 2004 IEEE 59th; 06/2004
ABSTRACT: In this paper, we present a novel synchronization scheme for multimedia applications that can provide high quality playback performance in emerging wireless environments. Because of the increasing size of global networks and the integration of wireless services, end-to-end delays are gradually increasing. Under these new conditions the existing synchronization schemes based on feedback, can no longer provide the necessary synchronization and thus quality needed for multimedia applications. We propose a novel synchronization scheme that requires no feedback, and which at the same time minimizes the required buffer space. We show that the proposed scheme performs very well, while significantly reducing communication overhead due to the control scheme by removing the feedback loop. Moreover, as its minimal buffer requirements are independent of network delays, the proposed scheme is inherently scalable for the evolving wireless environments.
Vehicular Technology Conference, 2003. VTC 2003-Fall. 2003 IEEE 58th; 11/2003
ABSTRACT: In this paper we study the inter- and intra- steam synchronization control schemes in networks with large end-to-end delays. We analyze pros and cons of different types of control schemes, focusing on the performance criterion of minimizing the jitter variance with the media stream for intra-stream synchronization, considered the most important objective. The minimum variance controller model is introduced to analyze the feasibility of feedback control schemes to achieve this objective. Due to the constraints of the distributed multimedia system model, especially when the average end-to-end network delay d<sub>avg</sub> is large (compared to the duration of the playback of the full buffer, 2Δ), we show that within a single stream (at least the master stream), the playback performance of any kind of feedback control scheme cannot be better than the uncontrolled system. Therefore, we can remove the feedback loop to reduce communication overheads and computation complexity. Given the increasing size of global networks and the integration of wireless services the delays are gradually increasing, making this result of increased importance in the emerging networking paradigms.
Communications, 2003. ICC '03. IEEE International Conference on; 06/2003
ABSTRACT: We solve the exact output tracking problem for square nonlinear
discrete-time systems through stable inversion. This stable inversion
problem involves finding a bounded solution of a dynamical system driven
by a bounded signal. Such a solution is given by a difference
representation formula, whose existence, under the proper assumptions,
can be proven using a Picard process
IEEE Transactions on Automatic Control 07/2000; · 2.11 Impact Factor
ABSTRACT: In this paper we build an observer for a nonlinear system which is
engaged in output tracking. In response to a desired bounded output
trajectory, there are now methods to compute the bounded inputs and
bounded states to achieve exact output tracking. To stabilize about the
desired state space trajectory (if states are not measured) we need
estimates of the error states, which are defined as the difference of
the actual states and the desired states. We can recover estimates for
the original states by adding the estimates of the error states to the
desired trajectory. We make use of the facts that the desired state
trajectory is bounded and that the error states are generally
Decision and Control, 1996., Proceedings of the 35th IEEE; 01/1997