1	PIES: Protocol Independent Energy Saving Algorithm Yasser Gadallah Thomas Kunz August 15, 2004
2	Presentation Outline Energy Consumption in Mobile Ad Hoc Networks Proposed Solution: PIES PIES Evaluation Conclusions
3	Energy Consumption In Mobile Ad Hoc Networks Many sources of energy consumption exist in mobile nodes e.g. screens, HDD, interfaces, etc Our focus is energy consumed in wireless interfaces Energy is consumed in sending, receiving, idle and sleep modes of operation It is crucial for energy conservation that we eliminate or reduce wasted idle energy consumption which constitutes a large portion of energy consumption is wireless interfaces To do this, ideally nodes need to be put to sleep when idle to save this wasted energy
4	Proposed Solution: PIES - Goals Fair energy conservation: all nodes are treated equally Complements existing routing algorithms from energy conservation perspective Functionality that is independent of the underlying routing protocol Little or no impact on network operation – no additional major traffic
5	Proposed Solution: PIES - Characteristics Achieves fair energy conservation by: Putting nodes to sleep for equal time periods Providing the routing algorithm with energy threshold info to help it make fair energy conscious decisions Fully distributed algorithm – functionality does not depend on a node or a set of nodes Modular nature – easily integrated with existing routing algorithms Ability to determine with certainty neighbors’ sleep state Configurable in such a way that introduces no additional traffic to the network
6	PIES Evaluation – Energy Performance
7	PIES Evaluation – Packet Delivery Performance
8	PIES Evaluation – Summary PIES introduces energy savings of about 50% for the conditions we used It also extends network lifetime by about 70% This is done while achieving higher fairness than in the case of the routing protocol alone The resulting PDR is comparable to that of the routing protocol alone Packet delivery latency increases with the increase of the ST/WT ratio
9	Conclusions Energy is the most scarce resource for the functionality of mobile ad hoc networks PIES is a solution that achieves fair energy conservation and works with existing routing algorithms Simulations show that PIES achieves energy savings of about 50% and extends network lifetime by about 70%
10	Backup Slides
11	Ad Hoc Network Energy Management Issues – Energy Conservation Challenges Energy conservation mechanisms should not add significant energy consumption or traffic demands on the network It should be able to answer the following questions: When to put the node to sleep without knowledge of traffic patterns? When to wake the node up? What is the sleep state of a neighbor? If the neighbor is asleep when will it wake up? There should be no major effect on the normal network functionality (e.g. no major traffic loss)
12	Energy-Efficient Schemes - Issues Most routing energy-efficient schemes ignored idle energy consumption in their analysis Non-routing energy-efficient schemes suffered one or more of the following issues: No attention given to energy fairness among network nodes Unrealistic assumptions e.g. traffic nodes do not forward traffic Algorithm design tied to a specific underlying routing strategy Large impact on the functionality of the underlying routing scheme No interaction with underlying routing protocol which results in unnecessary loss of traffic Heuristic methodology in determining neighbors’ sleep state
13	Proposed Solution: PIES - Operation