Chapter 14: Energy Efficient Next Generation Communications

1. HANDBOOK ON GREEN INFORMATION AND COMMUNICATION SYSTEMS Chapter 14: Energy Efficient Next Generation Communications 1Jason B. Ernst 1University of Guelph, Canada

2. Outline • Introduction • Applications and Networks • Architectures, Techniques and Protocols • Trends, Limitations, Challenges and Open Problems • Conclusions & Future Directions

3. Introduction • Motivation for Green Communications • Energy as a Resource • Limited Spectrum available • Unlicensed spectrum saturated (802.11abgn Wi-Fi, Bluetooth, Zigbee, Microwave Ovens, Cordless Telephones …etc.) • Interference between competing technologies • Particularly troublesome in dense network deployments • “Cognitive Radio” techniques to use licensed spectrum on demand when not in use • Limits of information transmission – “Shannon Limit”

4. Introduction • Motivation for Green Communications • Handheld & Battery Powered Devices • Batteries are heavy & expensive component • Reducing energy consumption allows smaller or cheaper power or increased lifetime • Many devices have multiple technologies (Wi-Fi, Bluetooth, Cellular), so choosing the best technology to reduce consumption in particular cases is important • Potential to extend overall lifetime of the network (in WSN or MANETs)

5. Introduction • Motivation for Green Communications • Potential Health Effects • Recent studies showing evidence for and against the health effects on humans of wireless exposure • Are models for exposure accurate? • Are standards for exposure keeping people safe? • Reduced Interference & Contention • Interference and contention implemented poorly can be inefficient • Collisions require retransmissions, using more energy • As the number of hops increase, without caching, the retransmissions become cumulative and successful communication becomes more unlikely

6. Applications and Networks • Wireless Sensor (WSN) & Mobile Ad Hoc Networks (MANETs) • Sensor nodes are often dropped randomly • MANETs can also have random topologies and node placement • Both WSN and MANETs cannot rely on infrastructure • Geographically may be dense or sparse in some areas • Nodes are often battery powered, have low processing and memory capability • WSN focuses on transporting data to collection points • MANET traffic may travel to and from Internet or between peers

7. Applications and Networks • Heterogeneous Wireless Networks (HWNs) • “Convergence Networks” • Use multiple technologies, ex: Bluetooth, Wi-Fi, 3G, 4G, Satellite • Each individual technology is optimized without considering inter-operability • Traffic may have the ability to select a more energy efficient route by selecting which technology to transmit over when several are available

8. Applications and Networks • Delay Tolerant Networks • Useful when network partitioning is possible (WSNs or MANETS

9. Applications and Networks • Delay Tolerant Networks • Often employs caching at individual hops so retransmission is not required across all hops, only at those where the data was lost • Can also employ scheduled transmission if some nodes appear on a predictable timeline (ex: orbits, bus schedules etc.)

10. Applications and Networks • Interplanetary & Intergalactic Networks • Communication in space has several challenges • Line of sight is often interrupted (planets and orbits) • There are often not many links and routes to choose from • Energy is limited, often solar powered • Weight is limited (rules out large batteries, and generally limits how much communication equipment and processing may be done) • Some traffic must be reliable (firmware updates, orbit corrections) • Other traffic may not be reliable (streaming video)

11. Applications and Networks • Alternative Energy Networks • These networks are not powered by a typical electrical grid • Solar, wind, tidal, etc. • Power may come in bursts, may or may not be stored in batteries • Communication may be intermittent when power is unreliable

12. Applications and Networks • Military, Emergency and Disaster Scenarios • Communication networks for these applications may be integrated with aerial drones, uavs, rovers, trained dogs, soldiers, medics, police offers, heavy equipment (tanks, planes etc) which require light and compact devices for mobilty • Again cannot rely on electrical grid • Must be secure and reliable

13. Applications and Networks • Developing Countries and Rural Access Networks • Internet access in these areas is extremely expensive • Also necessary to provide education, emergency services and communication • Should provide a tradeoff between performance and cost • In developing countries, a van physically drives between villages collecting emails which are sent opportunistically when the van returns to the city where infrastructure exists

14. Architectures, Techniques & Protocols • Topology Control • Many very different techniques to control the layout or arrangement of nodes in the network • Shut some nodes off completely periodically • Using a variety of cell sizes (smaller cells in dense areas so that fewer nodes are handled by one access point) • Adjust power levels dynamically to reduce neighbor connections (making routing choices easier)

15. Architectures, Techniques & Protocols • Power Control • Related to topology control, but not always done to control the layout of the network • May also be used to reduce destructive interference • Can be performed in conjunction with transmission schedules and compatibility matrices (used to determine which nodes may transmit at once without interfering – due to distance)

16. Architectures, Techniques & Protocols • Repeater and Relay Nodes • Can be used to reduce the distance required for long hops, reducing the transmission power • Especially useful in networks where battery life should be conserved • Similarly, special indoor antennas can be used within buildings so that mobile devices can use lower power when inside

17. Architectures, Techniques & Protocols • Caching, Clustering and Data Aggregation • Similar to repeater and relay nodes, cluster heads can be used to perform similar functionality • Nodes elect cluster heads within a nearby area and send all traffic through this node • The cluster head can then transmit several nodes’ traffic in burst

18. Architectures, Techniques & Protocols • Caching, Clustering and Data Aggregation

19. Architectures, Techniques & Protocols • Caching, Clustering and Data Aggregation • This reduces long hops to the destination, and avoids contention – reducing retransmissions • Techniques can also be used to switch cluster heads in the event the current cluster head is low on battery to keep the network “alive” longer • Particularly useful in WSN and MANETs

20. Architectures, Techniques & Protocols • Energy-Aware Handover Techniques • Traditional handover focuses on horizontal handovers (switching AP due to mobility) • With modern devices, vertical handover must also be considered (switching access technology due to mobility or changing network conditions) • Choice becomes more complex since this is a cost and overhead associated with handover • Using the correct information such as capacity, transmission energy, congestion, battery life can lead to improved efficiency

21. Architectures, Techniques & Protocols • Access Point Selection • Typically based solely on signal strength when choosing between multiple Aps • More recent solutions propose considering congestion level, capacity etc. • It is also possible to avoid using the infrastructure and form small ad hoc networks that relay eventually back to an AP

22. Architectures, Techniques & Protocols • Energy-Efficient Routing Algorithms • Particularly effective in WSNs • Consider a “data-centric” approach rather than “destination-centric” approach • Query based routing, for example: • “Give me periodic reports about animal location in region A every t seconds”

23. Architectures, Techniques & Protocols • Energy-Efficient Transport Protocols • In some application it may be worthwhile to tradeoff reliability for energy-efficiency • One way to encourage green communication is to prioritize traffic from other “green networks” who make energy efficiency a priority • Another way is to encourage protocols which attempt to reduce re-transmissions, ie avoid protocols like UDP which may cause many retransmission by flooding into the network without any feedback

24. Architectures, Techniques & Protocols • Energy-Efficient Transport Protocols • Removing the end-to-end requirement of traditional protocols and use a store-and-forward approach

25. Architectures, Techniques & Protocols • Energy-Efficient Medium Access Protocols • When combined with certain routing approaches such as Directed Diffusion (data-centric) it may be beneficial to abandon using automatic repeat requests (ARQ) in MAC protocols and assume that reliability is handled at higher layers • When links are periodically poor, using ARQ will reduce higher layer retransmissions (ie retransmission across every hop) • Trade-off between each decision, dynamically switch between strategies?

26. Trends, Limitations, Challenges & Open Problems • Integration of specific solutions for targeted applications, networks and architectures into a complete green strategy / framework • Joint optimization across existing wireless technologies • Existing networks are optimized without interoperability in mind • Need to optimize with respect to interaction for converging network solutions • Standardization

27. Conclusions & Future Directions • Apply known green techniques from specific networks and applications to more general networks • Combine approaches where possible, for ex) • Data centric routing + link layer ARQ disabled • Extend horizontal handover techniques to vertical • When multiple access technologies are available, make choice based on: • Capacity, Congestion level, Energy requirements etc. so that retransmission and contention are reduced

28. Thanks for your attention!