Md. Farhad Hossain , Kumudu S. Munasinghe and Abbas Jamalipour

1. A Protocooperation-based Sleep-Wake Architecturefor Next Generation Green Cellular Access Networks Md. FarhadHossain, Kumudu S. Munasinghe and Abbas Jamalipour Presented by: Yasser Mohammed

2. Motivations and Results • Over 80% of the energy in cellular networks is consumed at the access networks. In order to reduce the carbon footprint generated by cellular networks a novel energy efficient cellular access network architecture is proposed. With the adoption of the wake-up technology, base transceiver stations (BTSs) are made to protocooperatewith each other to achieve higher energy efficiency within a cellular access network. • Using the proposed network architecture and sleep-awake algorithm, energy in the order of 73% can be saved in individual base stations.

3. Outline • Section 2: Literature SurveyMotivation behind providing energy efficient access networks and the proposed system • Section 3: Ecological Balance and SustainabilityHow concepts of ecological balance can be applied to today’s homogeneous and future heterogeneous networks to develop a sustainable network infrastructure • Section 4: Proposed ArchitectureExamine the current cellular access networks. Propose the new network architecture and wake-up enabled base stations and examine their benefits • Section 5: Proposed AlgorithmAnalyze the Sleep-Wake (SLAKE) algorithm • Section 6: Simulation and Results

4. Background Sections 2 and 3

5. The radio access networks consumed 80% of the total energy consumed by cellular network infrastructure • More than 50% of cell-site operating expenditure is spent to power up base stations • Some suggested methods to save energy in cellular networks • shutting down some BTSs during low traffic time • high efficiency transmission scheduling • decreasing the size of the cells • design of improved BTSs which requires less cooling • sharing BTSs among different operators • using sectored cells • using renewable energy sources

6. Wake up technology has shown great potential in other fields of applications such as Personal Computers, Wi-Fi, Wireless Sensor Networks and Ethernet • Factors to consider when applying similar wake-up schemes in cellular networks: • highly dynamic nature of the network due to roaming users between BTSs • unpredictable mobility pattern • channel holding times of users • maintaining QoS of the services • wide coverage area per BTS • limited BTS capacity • time variant channel characteristics

7. Protocooperation is one of the ecological interactions by which two interacting species gain benefit through cooperation. However, this type of cooperation is not compulsory for the survival of any of these two species. • Different types of interactions among large number of homogeneous and heterogeneous components in an ecosystem play a key role in maintaining its sustainability. • NGMN is considered as an ecosystem, then different networks connected to the NGMN can be thought to be analogous to different components of the Earth’s heterogeneous ecosystem. • If communication networks are developed based on the principle of sustainability of the ecosystem those networks would be also self-sustainable.

8. How they obtained their results… Proposed architecture and algorithm

9. Sleep-Wake up enabled base stations • Base stations have 3 modes of operation: active, sleep and OFF • Active mode:BS is fully functional and both transmission and reception continue as normal • Sleep mode:a ‘wake-up module’ is located at each BS which will remain active to ‘sense’ or ‘hear’ any request from other BSs to switch in to active mode in the event of higher traffic arrival rates or the sudden failure of a neighbouring BS • OFF mode:The BTSs will remain disconnected from the power supply.

10. Protocooperative Cellular Access Networks • Cooperation happens through: • traffic sharing • bandwidth sharing • intelligent handovers • exchange of information on channel conditions • network loading

11. SLAKE algorithm

12. Facts, Figures and Numbers Simulation and results

13. Simulation Setup • MATLAB simulations • 180 overlapping macro cells having equal capacity of circular shape • Function used to generate traffic pattern at each BS cell: • Transmit power of carrier frequencies of acceptor BSs is programmable, power increment will be allowed if necessary • Cells are sectored and hence the BSs are equipped with directional antennas such that any power increment can be accomplished only in the specified direction for eliminating any deteriorating interference problems

14. The simulation time represents 24 hours of a day, which has been performed for three different cases: • similartraffic pattern for each BS, i.e., peak time and off-peak time coincides in all the BTSs • moderate variations in traffic patterns, i.e., peak time and off-peak time varies moderately among the BSs • relatively high variations in traffic patterns, i.e., peak time and off-peak time varies relatively high among the BSs.

15. BS saves 72.9% energy consumption in comparison to always active operating mode. • Certain period (14:00 hrs to 18:30 hrs) of its active time, it carries more traffic than it would have been carried out in an ‘Always-ON’ cellular system

17. Average sleeping time for the three scenarios is 10.2, 9.7 and 7.2 hours respectively and hence, energy savings are 42.5%, 41.0% and 30% respectively.

18. Take Away Points • BSs are equipped with more intelligence enabling them to make decisions by cooperating with each other for reducing energy consumption. • Proposed architecture is distributed, adaptive and autonomous. • This architecture is suitable for both homogeneous networks as well as next generation heterogeneous networks. • Proposed architecture and algorithm can save a substantial amount of energy at the access networks, which is very important for achieving a self- sustainable green communications future.