Wednesday, April 30, 2008

Adaptative Optimization of IEEE 802.11 DCF Based on Bayesian Estimation of the Number of Competing Terminals

A. López Toledo, T. Vercauterenm and X. Wang
Sequential Monte Carlo methods combined with a Bayesian approach is use to estimate the number of competing terminals to optimizw DCF parameters.

Monday, April 28, 2008

Voice Capacity Analysis of WLAN with Channel Access Prioritizing Mechanism

X. Ling,Y.Xeng,W. Shen, J. Mark

Most of the research in WLAN performance departs from the assumption of saturated stations. This was my motivation to write the WONS paper. In this article, the issue is solved in an elegant way. As a plus, they are capable to take into account different classes for QoS.

The article itself is divided in two parts. The first one studies the competition for the channel between VoIP and saturated sources in adhoc networks. The second part deals with the AP bottleneck problem when infrastructure operation. IMHO there is no tight relationship between the two parts of the articles, because the focus is placed in different problems.

Nevertheless, the article is well written, and will be an inspirations (and references) sources when writing my own articles.

Tuesday, April 15, 2008

C-PRMA: A Centralized Packet Reservation Multiple Access for Local Wireless Communications

G. Bianchi, F. Borgonovo, L.Fratta, L. Musumeci, M.Zorzi

Centralized PRMA is an extension of PRMA that includes a centralized scheduler that handles traffic hetereogeneity. IMHO the complexity is overwhelming.

The paper will be useful because its thorough introduction.

Thursday, April 3, 2008

T ree Algorithms for Packet Broadcast Channels

JOHN I. CAPETANAKIS

This paper introduces the concept of tree algorithms. The address space is divided in two halves. In two successive slots, the one is devoted to the first halve and the second to the second halve. If there is a collision, the address spaces is subsequently divided in two. Again the stations belonging to the first part transmit in the first slot while the stations included in the second part transmit in the second slot. This algorithm is repeated until all the stations have transmittet their packets.

Even though the proposal is theoretically sound and leads to throughput of 0.3-0.4 packets per slot, it probably present practical disadvantages. It was proposed in 1979 and we have not seen any implementation yet.

Wednesday, April 2, 2008

Design and Analysis of Cross-Layer Tree Algorithms for Wireless Random Access

Xin Wang, Member, IEEE, Yingqun Yu, Student Member, IEEE, and Georgios B. Giannakis, Fellow, IEEE

The reading of this paper resulted in many additional taks to do:
- There is a formula that is clearly wrong. Must find out what's goin' on
- It talks about tree algorithms for contention. What are this TA exactly?
- In the jitel paper, I didn't explain how to calculate the efficiency for the BEB case.
- The authors neglect the slots different lengths. Why? Is it because in 802.16 all the slots have equal length?
- Should I include this paper in the related work?