System Energy Efficiency Lab
Home People Research Publications Sponsors Contacts
Energy-efficiency in WLANs
PM in heterogeneous environment
Server controlled PM
WLAN scheduling for speech recognition

Energy-efficient wireless communication

Today’s wireless networks are highly heterogeneous with diverse range and QoS. The maintenance of a wireless link by a mobile device requires support of multiple network interfaces. Since the battery lifetime is limited, power management of their interfaces without any significant degradation in performance has become essential. In this page we present different approaches to achieve energy-efficient wireless communication.


Power management in heterogeneous wireless environments

We developed an integrated approach for the management of power and performance of mobile devices in heterogeneous wireless environments. Our policy decides what wireless network interface (WNIC) to employ for a given application and how to optimize the WNIC usage.

This decision is governed by:

  • the current power
  • performance needs of the system.

The policy dynamically switches between interfaces during program execution if data communication requirements and/or network conditions change.


For the verification of our power and performance management algorithm, we have experimentally characterized Bluetooth and 802.11b wireless interfaces. We implemented our policy on HP’s IPAQ portable device that is communicating with HP’s HotSpot server. The figure below shows the obtained results using various typical applications.


The results show that our policy offers a large improvement in power savings as compared to singly using 802.11b or Bluetooth while enhancing performance.


Server Controlled Power Management on wireless portable devices

This is a new power management technique aimed at increasing the energy efficiency of client-server multimedia applications running on wireless portable devices. We focus on reducing the energy consumption of the wireless network interface of the client by allowing the remote server to control the power con guration of the network card depending on the workload.In particular, we exploit server knowledge of the workload to perform an energy-efficient traffic reshaping, without compromising on the quality of service.

Our methodology exploits server knowledge of the workload, traffic conditions and feedback information from the client in order to minimize power consumption. The figure below shows theserver controlled power management architecture.

Two different power managers are needed: one running on the client (Client PM) and the other on the server (Server PM). The two PMs exchange power control information through a dedicated TCP connection. The following actions are performed:

  • The server decides when to enable the 802.11b PM. For example, in very light traffic conditions, the 802.11b PM might be used instead of a switch-off policy.
  • The server PM also schedules transmissions to the client in bursts, in order to compensate for the client performance and energy overheads during transitions between on and off states.
  • The client WLAN card is switched off once the server has sent a burst of data designed to keep the application busy till the next communication burst


We tested our methodology on the SmartBadge IV wearable device running an MPEG4 streaming video application.


Using our technique we measured energy savings of more than 67% compared to no power management being used on the WLAN interface. In addition, we save as much as 50% of energy with respect to the standard 802.11b power management. All of the energy savings are obtained with no performance loss on the video playback.


WLAN Scheduling to reduce the energy consumption of a distributed speech recognition front-end

In distributed speech recognition, speech features are computed on a mobile device, compressed, and sent to a server that performs the computationally intensive search for the most likely word sequence. A big challenge in designing a distributed speech recognition system is minimizing the energy consumption on the mobile device.

We consider quality-of-service tradeoffs including compression ratio and overall system latency and present a wireless LAN scheduling algorithm to minimize the energy consumption of a distributed speech recognition front-end on a mobile device.

Synchronous Scheduling of the WLAN Interface

We power down the 802.11b interface when not in use. The figure below shows the timing of our scheduling algorithm.


The period, T, is determined by the number of speech frames sent in one packet. The transmission is synchronous such that every T seconds we will send that amount of compressed speech features. Using the proposed scheduling algorithm, the WLAN card will only be on during the shaded region in the figure above.


  • 802.11 - In 802.11, we can couple the energy savings from larger packet sizes with a power on/off scheduling algorithm to reduce the overall
    energy consumption by a factor of 5 over the 802.11 PM mode in heavy broadcast traffic conditions
  • Bluetooth - We compare the results of this power saving algorithm to the low-power mechanisms of Bluetooth. The lower overhead of Bluetooth allows for greater energy savings with a much lower delay of approximately 300ms.