Imagine a scenario in which STAs from different providers contend for the bandwith of a single AP. Our reasoning is that the BE downlink traffic will dominate, since P2P and other symmetrical BE applications are not so popular as web browsing or media downloading in wifi scenarions.
802.11e provides us with the tools to control the uplink traffic, by assigning different EDCA parameters to each contending STA. And the one-million-dollar-question is... "Is this uplink control enough to control downlink flows as well".
TCP might help us in that direction, but we will need to simulate to find out... Teach me NS2!
Thursday, March 29, 2007
Using the 802.11e EDCF to Achieve TCP Upload Fairness over WLAN Links
D.J. Leith, P.Clifford
It studies the combined contention effects of TCP and DCF on uploading stations. NS2 simulations show a lockout problem that can be solved by giving preference to ack packets.
The authors also explain that in occasions in which the wireless hop is the bottleneck, the TCP dynamics are different from the traditional wired case.
It studies the combined contention effects of TCP and DCF on uploading stations. NS2 simulations show a lockout problem that can be solved by giving preference to ack packets.
The authors also explain that in occasions in which the wireless hop is the bottleneck, the TCP dynamics are different from the traditional wired case.
Tuesday, March 27, 2007
Online Association Policies in IEEE 802.11 WLANs
Gaurav S. Kasbekar, Joy Kuri and Pavan Nuggehalli
Oops I did it again, I read a paper that i cannot understand because I lack the mathematical background. It finds the optimal heuristic for a AP to chose to which AP to attach, and proves that its better that other policies.
There is a reference I want to read:
[6] A. Kumar, E. Altman, D. Miorandi, and M. Goyal, “New insights from a
fixed point analysis of single cell IEEE 802.11 WLANs,” in Proceedings
of the IEEE Infocom, 2005.
Oops I did it again, I read a paper that i cannot understand because I lack the mathematical background. It finds the optimal heuristic for a AP to chose to which AP to attach, and proves that its better that other policies.
There is a reference I want to read:
[6] A. Kumar, E. Altman, D. Miorandi, and M. Goyal, “New insights from a
fixed point analysis of single cell IEEE 802.11 WLANs,” in Proceedings
of the IEEE Infocom, 2005.
Tuesday, March 20, 2007
joining forces
After reading the last paper it appears clear to me that it is extremely difficult to write a good paper alone. The bibliography is so extense and the concepts so many, that team work is necessary to obtain medium or good results.
A Cross-Layer Approach for WLAN Voice Capacity Planning
Yu Cheng,Xinhua Ling,Wei Song, Lin X. Cai, Weihua Zhuang, Xuemin Shen
Interesting ingredients. A DCF-EDCA model with some similitudes with Boris', the idea of multiplexing downlink flows at the AP, my idea of dropping packets instead of retransmitting. Worth a second read...
Interesting ingredients. A DCF-EDCA model with some similitudes with Boris', the idea of multiplexing downlink flows at the AP, my idea of dropping packets instead of retransmitting. Worth a second read...
Hotspot Traffic Statistics and Throughput Models for Several Applications
Chen Na, Jeremy K. Chen, and Theodore S. Rappaport
Traffic analysis, inbound/outbound, throughput/SNR, blah blah... I'm more interested in a multiple user scenario, and this paper offers a reference...
[13] "Throughput measurements and models of Public IEEE 802.11b Wireless Local Area Networks". However, being the paper from the same authors, I don't fell like reading it. Not now.
Traffic analysis, inbound/outbound, throughput/SNR, blah blah... I'm more interested in a multiple user scenario, and this paper offers a reference...
[13] "Throughput measurements and models of Public IEEE 802.11b Wireless Local Area Networks". However, being the paper from the same authors, I don't fell like reading it. Not now.
Characterizing the IEEE 802.11 Traffic: The Wireless Side
Jihwang Yeo, Moustafa Youssef, Ashok Agrawala
First I have to say that I have not read the complete article. The appendices are so long that maybe it contains 30 pages alltogether. It performs a study listening the wireless medium and the conclusion is that less-than-optimal performance comes from less-than-optimal implementation of the protocol. Control frames are unnecessarily retransmitted and the selection of the bitrate is not in accordance of the actual SNR conditions. It is good to know but, from my point of view, in does not affect our research directly.
First I have to say that I have not read the complete article. The appendices are so long that maybe it contains 30 pages alltogether. It performs a study listening the wireless medium and the conclusion is that less-than-optimal performance comes from less-than-optimal implementation of the protocol. Control frames are unnecessarily retransmitted and the selection of the bitrate is not in accordance of the actual SNR conditions. It is good to know but, from my point of view, in does not affect our research directly.
Sunday, March 18, 2007
On the Performance Characteristics of WLANS:Revisited
Anand Balachadran, Geoffrey M. Voelker, Paramvir Bahl, P. Venkal Rangan
I love this paper.It basically takes all the findings that simulation-guys have been writing tons of papers about and says: this is bullshit, it has nothing to do with real world.
Now, what are the most interesting contributions... first it says that TCP and DCF make good friends, specially when the wireless terminals are clients downloading from the net. In this situation, the download throughput remains constant, no matter how many STAs enter into the network (up to 100).
HOWEVER, when the wlan reaches the saturation point, jitter and delays rise suddenly. This is not good for MM :-)
And this is funny: the automatic rate-adaption in 802.11 actually worsens the performance of .11 . It goes like this: when congestion occurs and a STA loses two consecutive packets due to collisions, the station changes to a lower rate effectively increasing the congestion.
And my conclusions: The DCF-TCP is the perfect team for web browsing, and file downloading from the net. However, as soon as new applications appear such as MM streaming, VoIP, P2P,... dcf can be a disaster. Would .11e be powerful enough to alleviate this situation?
I love this paper.It basically takes all the findings that simulation-guys have been writing tons of papers about and says: this is bullshit, it has nothing to do with real world.
Now, what are the most interesting contributions... first it says that TCP and DCF make good friends, specially when the wireless terminals are clients downloading from the net. In this situation, the download throughput remains constant, no matter how many STAs enter into the network (up to 100).
HOWEVER, when the wlan reaches the saturation point, jitter and delays rise suddenly. This is not good for MM :-)
And this is funny: the automatic rate-adaption in 802.11 actually worsens the performance of .11 . It goes like this: when congestion occurs and a STA loses two consecutive packets due to collisions, the station changes to a lower rate effectively increasing the congestion.
And my conclusions: The DCF-TCP is the perfect team for web browsing, and file downloading from the net. However, as soon as new applications appear such as MM streaming, VoIP, P2P,... dcf can be a disaster. Would .11e be powerful enough to alleviate this situation?
UniWireless: a Distributed Open Access Network
Danilo Severina, Mauro Brunato, Alessandro Ordine, Luca Veltri
Why is this paper important? First because they work on the OAN concept ant present a paper in WMASH every year. Second, because they ignored our contributions and failed to reference us year after year.
Third, and least important, because they introduce SIP register as a new authentication mechanism. The idea is brilliant, and actually is a topic of my interest, but imho the execution is poor. They include new SIP headers and mechanism to solve pass authentication information from proxy to proxy. And my question is ... Why making it so complicate instead of using IMS' well-known and extensevely tested procedures?
Why is this paper important? First because they work on the OAN concept ant present a paper in WMASH every year. Second, because they ignored our contributions and failed to reference us year after year.
Third, and least important, because they introduce SIP register as a new authentication mechanism. The idea is brilliant, and actually is a topic of my interest, but imho the execution is poor. They include new SIP headers and mechanism to solve pass authentication information from proxy to proxy. And my question is ... Why making it so complicate instead of using IMS' well-known and extensevely tested procedures?
Friday, March 16, 2007
Secure Access to IP Multimedia Services Using Generic Bootstrapping Architecture (GBA) for 3G & Beyond Mobile Networks
Muhammad Sher Thomas Magedanz
What I want to know is how a UE accesses the Internet (for example, for web browsing) in a IMS environment. In this paper the bootstrap procedure that allows a ue to access a IMS services from a generic network. It is somehow related to what I wanted, but not what I wanted...
For me it is more or less clear how the ue accesses sip-based applications through the IMS proxies, but not how non-sip based Internet applications are used...
What I want to know is how a UE accesses the Internet (for example, for web browsing) in a IMS environment. In this paper the bootstrap procedure that allows a ue to access a IMS services from a generic network. It is somehow related to what I wanted, but not what I wanted...
For me it is more or less clear how the ue accesses sip-based applications through the IMS proxies, but not how non-sip based Internet applications are used...
Thursday, March 15, 2007
Comprehensive Analysis of the IEEE 802.11
PETER P. PHAM
It seems that I finally found what I was looking for. This articles might contain the formulas that I can use for my ideas.
It seems that I finally found what I was looking for. This articles might contain the formulas that I can use for my ideas.
Resource Control for the EDCA Mechanism in Multi-Rate IEEE 802.11e Networks
Vasilios A. Siris Costas Courcoubetis
The article uses pricing to obtain best utilization of the wireless channel using EDCA. I had already read some papers about using pricing in response to congestion.
However, I'm more interested in some of the references which can provide me with the formulas/approximations required for my next article.
The article uses pricing to obtain best utilization of the wireless channel using EDCA. I had already read some papers about using pricing in response to congestion.
However, I'm more interested in some of the references which can provide me with the formulas/approximations required for my next article.
Tuesday, March 13, 2007
Performance Analysis and Enhancements for IEEE 802.11e Wireless Networks
Qiang Ni, National University of Ireland
This paper is a *MUST*. The one you should begin with. Well-written, comprehensive,... all you need to know about .11e... specially if you expect to understand any of the others papers.
This paper is a *MUST*. The one you should begin with. Well-written, comprehensive,... all you need to know about .11e... specially if you expect to understand any of the others papers.
idea: multi-tiered CAC
It might seem trivial to you, dear reader, but it took me three weeks to realize what's my next pape r about.
This is my conclusion:
Step 1: Take any CAC
Step 2: Say that in addition to differentiate between VO and BE, we want to offer different QoS to paying and non-paying users.
Time to write?
This is my conclusion:
Step 1: Take any CAC
Step 2: Say that in addition to differentiate between VO and BE, we want to offer different QoS to paying and non-paying users.
Time to write?
idea: Using neighbour cells population together with to compute roaming probability
The probability that a call roams into our cell highly depends on the number of ongoing calls in the neighbouring cells. This information could be used to compute guard channels etc.
Extending End-to-End QoS to WiFi based WLAN
http://www.cswl.com/whitepapers/qos-wireless-lan.html
Plain and clear. QoS in WLAN and interaction with SIP. It does not specify the details. Which could be our contribution? A problem to be solved (inmho) is the mapping from a codec specified in SDP to a TSPEC, and then (taking into account current traffic conditions) the conversion to EDCA parameters.
Plain and clear. QoS in WLAN and interaction with SIP. It does not specify the details. Which could be our contribution? A problem to be solved (inmho) is the mapping from a codec specified in SDP to a TSPEC, and then (taking into account current traffic conditions) the conversion to EDCA parameters.
Monday, March 12, 2007
WLAN Manager (WM): a Device for Performance Management of a WLAN
Malati Hegde† , S.V.R Anand† , Anurag Kumar† , and Joy Kuri‡
The most useful idea that we can obtain from this article is that the access point controller defined by CAPWAP can be the actual policy enforcement point.
The details about how to obtain "fairnes" understood as giving more throughput to the users with higher rates by recdistributing the packets in the queues and pinging the STA to find out their rates seemed to me a bad idea. Additionally, there are typos. So I decided to read only journal papers from now on. I was surprised when I realized that this was a journal paper :-(
The most useful idea that we can obtain from this article is that the access point controller defined by CAPWAP can be the actual policy enforcement point.
The details about how to obtain "fairnes" understood as giving more throughput to the users with higher rates by recdistributing the packets in the queues and pinging the STA to find out their rates seemed to me a bad idea. Additionally, there are typos. So I decided to read only journal papers from now on. I was surprised when I realized that this was a journal paper :-(
Localizing the Internet, Five Ways Public Ownership Solves the U.S. Broadband Problem
Becca Vargo Dagget
An extensive study of the last mile bottleneck in the U.S. The author defents municipality owned access network to foster competence, assure network neutrality, universal service, higher incomes for the municipality and lower prices for the users.
It contains case studies and success stories.
An extensive study of the last mile bottleneck in the U.S. The author defents municipality owned access network to foster competence, assure network neutrality, universal service, higher incomes for the municipality and lower prices for the users.
It contains case studies and success stories.
Thursday, March 8, 2007
LocustWorld Mesh explained
Jon@LocustWorld.com
Similary to linux live CDs, LocustWorld provides a boot CD that converts the PC in the node of a mesh network. We sould try it and think how it could be useful to us.
Similary to linux live CDs, LocustWorld provides a boot CD that converts the PC in the node of a mesh network. We sould try it and think how it could be useful to us.
Wednesday, March 7, 2007
idea- dropping delayed voice packets at the MAC layer
It seems to me that the use of Access Categories in EDCA is unfair by definition. A user might decide to place all its traffic (even if is P2P file sharing ) in a high priority queue. That user will obtain higher bitrates and lower latency at the expense of the users that place each kind of traffic to the appropriate queue.
Additionally, delayed voice packet might be useless at the receptor. If a packet voice is delayed due to retransmission attempts, it will provoke the delay of the rest of the packets in the queue and could happen that all of them had to be dropped because the delay budget has been exceeded.
My proposal is that if a voice packet suffers a collision, it should be dropped and the contention window for the following packet should be doubled. With my proposal, the high priority queues will obtain higher bit rates and lower delays (thanks to shorter AIFS and smaller contention windows). On the other hand, lower priority queues will obtain higher reliability thanks to the retransmission mechanism.
Additionally, delayed voice packet might be useless at the receptor. If a packet voice is delayed due to retransmission attempts, it will provoke the delay of the rest of the packets in the queue and could happen that all of them had to be dropped because the delay budget has been exceeded.
My proposal is that if a voice packet suffers a collision, it should be dropped and the contention window for the following packet should be doubled. With my proposal, the high priority queues will obtain higher bit rates and lower delays (thanks to shorter AIFS and smaller contention windows). On the other hand, lower priority queues will obtain higher reliability thanks to the retransmission mechanism.
A performance study on the service integration in IEEE 802.11e wireless LAN
Frank Roijers, Hans van den Berg, Xiang Fan, Maria Fleuren
The goal of the article is to compare the influence of different EDCA parameters in the performance metrics of different fluxes by means of simulation.
There is a mistake in fig 3. According to this plot (Fig. 3, right), stations with a lower CWmin (7) obtain a lower throuhput than those with a higher CWmin (31).
The article offer a visual explanation of the effect of tuning the different EDCA parameters.
The goal of the article is to compare the influence of different EDCA parameters in the performance metrics of different fluxes by means of simulation.
There is a mistake in fig 3. According to this plot (Fig. 3, right), stations with a lower CWmin (7) obtain a lower throuhput than those with a higher CWmin (31).
The article offer a visual explanation of the effect of tuning the different EDCA parameters.
An Admission Control Heuristic for IEEE 802.11e Wireless LANs
Bechir Hamdaoui, Moncef Elaoud, Parameswaran Ramanathan
It proposes a CAC mechanism similar to the one that Cristina presented in one of the NeTS seminars. The different lays in the model that is used to calculate the delays suffered by the multimedia packets. It seems that the model together with the CAC mechanism successfully predicts the number of VoIP calls that an AP can support, compared to simultion.
It does not contemplate channel errors nor rate adaption.
I lack the mathematical knowledge to fully understand the model and therefore I should re-read the paper and check the different steps one-by-one.
It proposes a CAC mechanism similar to the one that Cristina presented in one of the NeTS seminars. The different lays in the model that is used to calculate the delays suffered by the multimedia packets. It seems that the model together with the CAC mechanism successfully predicts the number of VoIP calls that an AP can support, compared to simultion.
It does not contemplate channel errors nor rate adaption.
I lack the mathematical knowledge to fully understand the model and therefore I should re-read the paper and check the different steps one-by-one.
idea
It seems that most of the aspects of 11e and 11r have been studied. So let's focus on the multi-operator environment. Let's assume that two different operators share an access point broadcasting two different ssids. What's the impact of the traffic of one operator on the QoS of the other operator?
Tuesday, March 6, 2007
Performance Study of Fast BSS Transition using IEEE 802.11r
Sangeetha Bangolae Carol Bell Emily Qi
Her lays the answer to the previous post. The negotiation can be done prior to roaming. Additionally the STA can, can request roaming to different APs, and then choose one to roam.
Her lays the answer to the previous post. The negotiation can be done prior to roaming. Additionally the STA can, can request roaming to different APs, and then choose one to roam.
idea
What happens with QoS negotiation in 802.11r? Should it be negotiated *before* handover? What can we propose?
Integrating SIP and IEEE 802.11e to Support Handoff and Multi-grade QoS for VoIP Applications
Jen-Jee Chen Ling Lee Yu-Chee Tseng
Very complete! It addresses...
(1) Handoff
(2) Multigrade QoS
(3) Multirate
(4) Multiple codecs
(5) A bit of SIP
(6) Call Admission Control
(7) Resource re-Adjustment
What else can we add...
(1) Including location information to estimate the handoff probability (Ph).
(2) Using Boris-Model (Bandwidth-Orinented Resource Inferring System) to calculate the resource utilization of a new flux.
Very complete! It addresses...
(1) Handoff
(2) Multigrade QoS
(3) Multirate
(4) Multiple codecs
(5) A bit of SIP
(6) Call Admission Control
(7) Resource re-Adjustment
What else can we add...
(1) Including location information to estimate the handoff probability (Ph).
(2) Using Boris-Model (Bandwidth-Orinented Resource Inferring System) to calculate the resource utilization of a new flux.
Friday, March 2, 2007
Design of QoS and Admission Control for VoIP Services over IEEE 802.11e WLANs
Pei-Yeh Wu1 , Yu-Chee Tseng1 , and Hewitt Lee2
Very complete and extense (25 pags). A good place to begin with. Explains all the basic concepts. Introduction of tutorial nature. It makes a shy attempt to link SIP and EDCA. It addresses codec selection and Packitization Intervals.
Very complete and extense (25 pags). A good place to begin with. Explains all the basic concepts. Introduction of tutorial nature. It makes a shy attempt to link SIP and EDCA. It addresses codec selection and Packitization Intervals.
Thursday, March 1, 2007
New QoS Control Mechanism Based on Extensions to SIP for Access to UMTS Core Network via Different Kinds of Access Networks
Mehdi Mani, Student Member, IEEE ,Noël Crespi
Very interesting. Bad English. It addresses architectural issues, SIP integration, IMS... . It includes a testbed using WLAN. It lacks the use of EDCA as QoS mechanism. That could be our contribution.
Very interesting. Bad English. It addresses architectural issues, SIP integration, IMS... . It includes a testbed using WLAN. It lacks the use of EDCA as QoS mechanism. That could be our contribution.
Wireless LAN Resource Management Mechanism Guaranteeing Minimum Available Bandwidth for Real-time Communication
Toshihiko Tamura and Tadashi Ito
Well written paper. It addresses the codec adaptation in combination with SIP and 802.1X. It introduces AP-RMS (AP-Resource Management Service). However, it leaves as future work the task of considering rate adaptation, or EDCA parameter tuning. IMHO, their work can be very well complemented by Boris and Anna's solutions.
Well written paper. It addresses the codec adaptation in combination with SIP and 802.1X. It introduces AP-RMS (AP-Resource Management Service). However, it leaves as future work the task of considering rate adaptation, or EDCA parameter tuning. IMHO, their work can be very well complemented by Boris and Anna's solutions.
A Priority Control Method for High quality VoIP Systems over Wireless LAN
R. Matsukura, H. Ymauchi, S. Okuyama, M. Matsuda, M. Sumioka
Badly written, hard to read. The contributions are not clear. It might be of interest because addresses a topic related to our work.
Badly written, hard to read. The contributions are not clear. It might be of interest because addresses a topic related to our work.
Subscribe to:
Posts (Atom)