PDF Performance Analysis of IEEE 802.11 DCF Power Save Mode in IBSS - ERNET

Raju, faculty, MNC, IIT Guwahati for their help and encouragement in the initial stage of my thesis work. I am also thankful to my brother (bhai) and sister in law (Bhauja), uncle (dadu), grandmother (Maa), Baba, mother for their love and encouragement.

Abstract

List of Tables

List of Algorithms

Introduction

Motivation of the Research Work
Contributions of the Thesis

Performance Analysis of IEEE 802.11 PSM in IBSS for Saturated Traﬃcfor Saturated Traﬃc
Performance Analysis of IEEE 802.11 PSM in IBSS for Unsaturated Traﬃcfor Unsaturated Traﬃc

Organization of the Thesis

The IEEE 802.11 standard [2] specifies the medium access control (MAC) and physical (PHY) layer specifications for a WLAN device. IEEE 802.11 DCF PSM throughput in IBSS is analyzed using an analytical model of ATIM frame and data frame transmissions.

Background and Literature Survey

The IEEE 801 Distributed Coordination Function (DCF)

Carrier Sense Mechanism
Binary Exponential Backoﬀ

The IEEE 802.11 DCF in Power Save Mode (PSM)

Power Management in Infrastructure Network
Power Management in IBSS

Modeling of IEEE 802.11 DCF

Extensions of Bianchi’s [1] Model for Saturated Traﬃc ConditionCondition
Performance Modeling of IEEE 802.11 DCF in Unsat- urated Traﬃc Condition

Analytical Model of IEEE 802.11 DCF in PSM
Summary

There are a number of theoretical models that analyze the performance of the IEEE 802.11 DCF depending on different network assumptions. This model provides quality of service (QoS) performance and queuing behavior of the IEEE 802.11 DCF.

Figure 2.1: Inter Frame Spacing in IEEE 802.11

Modeling IEEE 802.11 Power Save Mode in IBSS for Saturated

Traﬃc

The IEEE 802.11 DCF in Power Save Mode

The probability of successful transmission of an ATIM frame has a major impact on a node's data frame transmission in IBSS PSM. Therefore, the probability of successful transmission of an ATIM frame is required for throughput calculation for IEEE 802.11 DCF PSM in IBSS. If the ATIM frame is not successfully sent after the third ATIM window, the data frame is discarded.

The stations that successfully send an ATIM frame within the ATIM window compete to send a data frame in the rest of the BI. If the ATIM frame is not successfully sent after three ATIM windows, the data frame is discarded.

Modeling and Analysis

Network Model Assumptions
System Model
Model Analysis
Analysis and Estimation of q a and q d
Saturation Throughput Analysis

Similarly, qd is proportional to the number of active stations in the data window. The eleventh equation indicates that the station increments the backoff stage and selects the backoff counter uniformly after an unsuccessful transmission of a data frame within the data window. It has been assumed in the model in Figure 3.1 that qa is the probability of reaching the end of an ATIM window and qd is the probability of reaching the end of a data window.

The probability value qd is proportional to the number of active stations in the data window and the data window size. For a fixed ATIM window size, the value of c depends on the size of the data window.

Figure 3.1: Markov model for data frame transmission in power save mode

Analytical Model for Delay Analysis

Dsucc(a) (k) =k×BI + ATIMsize (3.24) Assume that D(d)succ(i, b) is the delay to transmit a data frame successfully in the data window's ide back-off stage when the sum of the backoff values up to stage i are b. Since (Wi−1) is the maximum CW magnitude at the current stage, the average value of b is W i2. Psucc(a)(i, k) = Xki(1−pa)(1−qa) (3.28) where Xki is the probability that a station will attempt to send an ATIM frame in the last back-off step in the k. ATIM window.

Psucc(d)(i, b) is the probability that a data frame is sent at the ith stage, and b is the sum of the delay values up to the ith stage of delay. Assume that Psucc(d)(i) is the probability of successful transmission of a data frame in the ith offset phase of the data window, and pd(1−qd) is the probability that a collision occurs in the data window.

Analytical Model for Power Consumption

Analytical Model for Standard Deviation of Delay

Average time spent sending an ATIM frame in the ATIM window (Ttx/rxa), and . Average time spent sending a data frame in the data window (Ttx/rxd), after successful transmission of the ATIM frame in the ATIM window. A station can be in idle mode in the ATIM window or in the data window.

The station waits W time units before sending a frame and decrements the return counter after each slot time (assuming each slot is an empty slot). 5I didn't use conditional probability because this means that the ATIM/data frame transfer offset process ends at a certain offset level, given that an ATIM frame is successfully sent.

Model Validation and Performance Evaluation

Model Validation and Performance Evalua- tion

Analysis of the Proportional Constant c
Saturation Throughput
Delay Analysis
Power Consumption
Comparison between DCF with and without Power Save ModeSave Mode

The size of the beacon interval is changed in simulation to study the behavior of the model. The second observation is that the data window throughput is almost the same for different beacon interval sizes. With the increase in beacon interval size, the ATIM overhead decreases (since the ATIM window size is kept fixed).

It can be seen from the figure that as the size of the beacon interval increases, the average power consumption decreases. That is why, as the beacon interval size increases (with a fixed ATIM window size), the average delay increases, as shown in Table 3.2 and Figure 3.5.

Table 3.1: Parameters used in the simulation Parameter value

Summary

Modeling IEEE 802.11 IBSS

Power Save Mode for Unsaturated Traﬃc

Modeling and Analysis of IEEE 802.11 PSM for Unsaturated Traﬃc

Network Model Assumption
Modeling and Analysis
Model Analysis
Estimation of probability α, q a , q d
Throughput Analysis

The number of active stations participating in channel access during the data window is selected based on the probability of successful transmission of an ATIM frame. Let bi denote the conditional collision probability in the ATIM window and pd the same for the data window. The second equation shows the probability α that a station has a data frame in the buffer to transmit.

Let τd be the probability that a station transmits a data frame in a randomly selected slot in the data window. 4.17) kun′′is the number of active stations participating in contention in the data window.

Figure 4.1: Markov model for IEEE 802.11 IBSS PSM with unsaturated traﬃc

Analytical model for Average MAC Delay

Let Psucc(a)(i, k) be the probability that an ATIM frame is successfully sent at the ith offset stage of the kth ATIM window, and Pdrop(a) denotes the probability that an ATIM frame is dropped due to a retry. limit exceeded in last ATIM window. Let B(i) be the total value of the backlog up to the i-th backlog level, and B(i)max be the sum of the largest sizes of the competitive window up to the i-th backlog level. The value of Psucc(d)′(i, b) indicates the conditional probability that the data frame transfer offset process ends at the ith stage of the data window, with the total offset value b up to the ith offset stage, given that the data frame is successfully downloaded.

Here Pdrop(d) is the probability of dropping a data frame that exceeds the retry limit in the data window. Psucc(d)(i) ={pd(1−qd)}i{(1−pd)(1−qd)} (4.38) Let Pidle(d), Pcol(d) and Psucc(d) be the probability that a randomly selected slot in the data window is inactive, causes a collision, and results in a successful transmission, respectively.

Analytical Model for Average Power Consumption

Analytical Model for Average Power Consump- tion

Standard Deviation of Power Consumption

Similarly, let Psucc(d)(i) denote the probability of successfully transmitting a data frame in the ith backoff phase of the data window. The station is in the back-off phase, so it remains idle during the back-off period. The station has successfully sent an ATIM frame, so by default it should stay awake in the data window.

In the DATA window, the station remains idle after a successful download if all these conditions are met. me). Idle Power: A station can go into Idle mode if it neither transmits nor receives an ATIM frame in the ATIM window.

Model Validation and Performance Evaluation

Model Validation and Performance Evalua- tion

For the unsaturated case, Figure 4.4 and Figure 4.5 (results obtained from analytical model) show how the normalized throughput depends on the size of the network and the size of the BI. It can be seen from Figure 4.4 that when the size of BI is increased with a fixed ATIM window of 20 ms, the normalized network throughput increases. The impact of network size and BI size on power consumption is.

From the figure it can be observed that with the increase in the size of BI, the average energy consumption decreases. It is also observed that with the increase in the rate of data arrival, the amount of energy consumption decreases.

Summary

The performance of IEEE 802.11 IBSS PSM is also analyzed and compared to standard IEEE 802.11 IBSS without PSM. The analysis shows that there is a trade-off between throughput, MAC lag, and average power consumption. A dynamic and adaptive BI can provide better energy savings with little compromise on latency and throughput, which can open up several directions for future research.

Conclusions and scope of Future Work

Summary of Contributions

Analytical models for normalized network throughput, average MAC lag, and average power consumption in the IEEE 802.11 IBSS PSM are then presented. Using the proposed model, a comprehensive and insightful performance evaluation of IEEE 802.11 IBSS in system parameters in power saving mode is performed. It has also been found that the dynamic tuning of the system parameters has a major impact on achieving optimal performance in IEEE 802.11 IBSS in power saving mode.

Performance modeling of IEEE 802.11 PSM in IBSS for unsaturated traffic: In practice, wireless networks do not operate in saturated traffic conditions. The impact of data arrival rate, network size and the Beacon Interval size on the performance of the IEEE 802.11 DCF in PSM is also analyzed.

Scope of Future Work

The model can be extended to analyze the performance of M-group heterogeneous IEEE 802.11 DCF PSM in IBSS with different arrival rates, network sizes and beacon interval sizes in each group. The model can be further extended to develop an analytical model of IEEE 802.11 DCF power saving mode under saturated and unsaturated traffic conditions in multi-hop ad hoc networks. The problems that may arise in the IEEE 802.11 DCF power saving mode when operating in a multi-hop ad hoc network in the presence of hidden nodes can be investigated.

The performance of the IEEE 802.11 DCF in multi-hop ad hoc networks for different traffic loads, packet sizes, and carrier detection ranges can be analyzed. An analytical model of IEEE 802.11 DCF power saving mode in multi-hop ad hoc networks based on queuing theory can be designed to study the average throughput share between the different priority classes.

Jayasuriya, "Performance analysis of IEEE 802.11 DCF under limited load," in Proceedings of the Asia-Pacific Conference on Communications. Spasenovski, "Saturation throughput-delay analysis of IEEE 802.11 DCF in fading channel", in Proceedings of the International Conference on Communications, vol. Paparrizos, “Delay distribution analysis of IEEE 802.11 with variable packet length,” in Proceedings of the Vehicular Technology Conference.

Yang, "Throughput Modeling and Analysis of IEEE 802.11 DCF with Selfish Node", in Proceedings of the Conference on Global Telecommunications. Xiao, "Modeling IEEE 802.11 DCF System Dynamics", in Proceedings of the Wireless Communications and Networking Conference.

Publications Related to Thesis

Journals

Conference Proceedings

Other Publications of the Author

Brief Biography of the Author

Indian Institute of Technology Guwahati