[Paper Review] Analytical Models for Energy Consumption in Infrastructure WLAN STAs Carrying TCP Traffic

Pranav Agrawal, Anurag Kumar+5 more|ArXiv.org|Sep 21, 2009

Wireless Networks and Protocols11 references4 citations

TL;DR

This paper develops analytical models to estimate energy consumption in infrastructure WLAN stations (STAs) performing TCP-based file transfers, comparing Continuously Active Mode (CAM) and Power Save Mode (PSM). It shows that PSM outperforms CAM for short file transfers due to reduced idle power, but CAM is more efficient for long continuous downloads due to PSM's overhead, validated via NS-2 simulations and Markov chain modeling of STA state transitions.

ABSTRACT

We develop analytical models for estimating the energy spent by stations (STAs) in infrastructure WLANs when performing TCP controlled file downloads. We focus on the energy spent in radio communication when the STAs are in the Continuously Active Mode (CAM), or in the static Power Save Mode (PSM). Our approach is to develop accurate models for obtaining the fraction of times the STA radios spend in idling, receiving and transmitting. We discuss two traffic models for each mode of operation: (i) each STA performs one large file download, and (ii) the STAs perform short file transfers. We evaluate the rate of STA energy expenditure with long file downloads, and show that static PSM is worse than just using CAM. For short file downloads we compute the number of file downloads that can be completed with given battery capacity, and show that PSM performs better than CAM for this case. We provide a validation of our analytical models using the NS-2 simulator. In contrast to earlier work on analytical modeling of PSM, our models that capture the details of the interactions between the 802.11 MAC in PSM and certain aspects of TCP.

Motivation & Objective

• To model energy consumption in WLAN STAs during TCP-controlled file downloads, focusing on radio energy in CAM and PSM modes.
• To analyze the impact of file size (long vs. short) on energy efficiency in CAM and PSM modes.
• To develop analytical models using Markov chains and Processor Sharing (PS) models to estimate state occupancy (idle, receive, transmit) and energy expenditure.
• To validate the models using NS-2 simulations and compare analytical results with simulation outcomes.
• To identify performance trade-offs between PSM and CAM under varying traffic patterns and STA counts.

Proposed method

• Modeling STA radio state transitions (idle, receive, transmit) using a continuous-time Markov chain with state-dependent transition rates.
• Deriving stationary probabilities of system states using the theory of Markov Regenerative Processes (MRGP).
• Using a Processor Sharing (PS) model to approximate downlink throughput in short file transfer scenarios, with service rate derived from long file analysis.
• Formulating the expected energy expenditure as a function of state occupancy fractions and transmission parameters.
• Incorporating realistic PSM behavior: STAs avoid re-sending PS-POLL frames during waiting, reducing contention.
• Validating analytical results against NS-2 simulations using HTTP-like traffic with exponentially distributed file sizes and think times.

Experimental results

Research questions

• RQ1 How does energy consumption in WLAN STAs differ between CAM and PSM modes during long TCP file downloads?
• RQ2 What is the impact of short file transfers and think times on the energy efficiency of PSM versus CAM?
• RQ3 How does the number of associated STAs affect the performance and energy efficiency of PSM and CAM?
• RQ4 To what extent do contention and beacon frame overhead degrade PSM performance in high-density scenarios?
• RQ5 Can analytical models accurately predict STA energy consumption for TCP traffic in both CAM and PSM modes?

Key findings

• For long file downloads, CAM results in lower energy expenditure than PSM due to the overhead of PS-POLL frame contention and wake-up delays in PSM.
• For short file transfers with think time, PSM enables significantly more file downloads per unit battery capacity than CAM, as STAs spend more time in low-power sleep states.
• The analytical model's predicted energy expenditure and file transfer counts closely match NS-2 simulation results, validating the model's accuracy.
• With increasing numbers of STAs, PSM efficiency degrades due to increased overhearing of frames destined for other STAs, reducing the benefit of sleep states.
• The stationary probability of multiple STAs being active simultaneously is non-negligible, especially at N=8, indicating a key bottleneck in PSM scalability.
• Increasing the beacon interval can improve PSM efficiency by reducing wake-up frequency, but at the cost of increased delay.