GREEN Solutions for Wireless Systems WHITE paper

www.hcltech.com GREEN Solutions for Wireless Systems WHITE paper

TABLE OF CONTENTS Introduction RAN Optimizations Energy Efficient Base Station (NodeB/ enodeb) Energy Efficient Controller RAN Optimizations Transport / Backhaul / Cabinet Optimizations BBU Hotelling Reduce the Heating cost/ Auto Cooling CPU throttling / Dynamic Frequency Scaling Power Transmission Use tower mounted Radio to eliminate feeder loss Conclusions References 3 5 5 7 8 9 9 9 9 10 10 10 11

Introduction In our view, the idea for GREEN is: Getting a Reliable Energy Efficient Network. The objective of such GREEN initiatives is to reduce the energy consumption of the wireless communication network, and thereby achieve Operation Expense (OPEX) savings and CO2 emission reduction. Rapid growth in the number of smartphones and the increase in number of connected equipments i.e. Machine-2-Machine (M2M) are going to further increase the need for more base stations. For example, a wireless operator in China has seen a 81x times growth in the user base in last 5 years, while the annual power consumption of its 2G infrastructure is estimated to be 80GWH [1]. In a wireless communication system such as 3G/4G, RAN part accounts for almost 1/2 of the total power consumption, and the base stations account for about 2/3 of the RAN power consumption. The rough divide of the power among network equipments is as shown in Fig.1 [1]: Power consumption Ran 51% Core 9% Ran 51% Transmission 12% Aggregators 5% FTTH 5% Fig. 1: Power consumption in Wireless Systems Within a base station (enodeb), we have radio and digital hardware. The radio components usually have very low power efficiency and contribute to maximum power utilization. A 60W RRH could consume about 300W power. Moreover, there is only 10% difference between typical and maximum power consumption of a RRH. Hence, for reducing power consumption, it is important to make efficient use of the radio component. The Fig.2 below shows power breakup across components in base station.

power consumption in enodeb 900 800 700 600 500 400 300 200 100 0 Coolingsystem Controller Modem Rrhs Watt Fig. 2: Power utilization per component in base station (enodeb) Even though the digital hardware, which runs the Layer1 and Call processing functionality consumes less power when compared to radio equipments, making efficient use of the available hardware resources would contribute in improving the Total Cost Ownership (TCOs) for the operator and would also reduce the power consumption. Lack of continuous supply of grid electricity in the highly populated and large wireless markets such as India, pushes the wireless operator to use a diesel generator to power-up the sites. The diesel electricity is not only expensive in terms of cost and CO2 emission but also needs human intervention. Some base stations could be located at remote places, where, getting access to the site and sending a person for maintaining the diesel generator could add up to significant operational expenses. GREEN provides architectural solutions for achieving energy efficiency. The main objective is to optimize the power consumption by update of architectural solutions, thereby reducing the wastes from the existing systems and increasing the efficiency. The Fig.3 below describes the overall architectural solutions choices for achieving GREEN. The solutions are spread across the network elements and mainly concentrate on achieving energy efficiency through decreasing the number of base stations required, reducing the hardware required to provide the service, decreasing the cooling requirement, and achieving higher spectral efficiency. The GREEN solution aims to increase the efficiency of the wireless system without compromising on the coverage, user peak data rate, and UE (User Equipment) battery life. The various aspects of the solution are discussed below. Energy usage: With improvements in design, keep on reducing the normal day-to-day waste & consumption. Sustainable: Keep running continuously and keep on reducing the current usage pattern & if possible, using the renewable resources.

Spectrum Efficiency: Making the information transfer over the air efficient, with more number of information Bits on a particular Bandwidth (per Hz). Traffic monitoring Cpu thrattling Bbu hotteling Higher layer modulation Smart antennas Load sharing across cells Carrier aggregation for higher data rates Coverage optimizations Lower frequency for higher caverage Multi-core cp us Extend temp range Power failure detection Use higher voltage to cover large distances Beam forming Alternate power resources Reduce cooling-air / convention ways Fig. 3: Proposals for achieving the overall GREEN impact in Wireless Systems Optimizations: To keep on innovating in the current systems while optimizing power consumption and providing the same system performance. Smart Antenna: Smart Antennas are AAA (Adaptive Array Antennas) with smart signal processing algorithms, which identify direction of arrival of the signal, and then calculates the beam-forming vectors, which would track and locate the antenna beam on the mobile/target. RAN Optimizations The RAN part accounts for more than 50% of the total power consumption in the Wireless Systems, so operators should try to keep on optimizing and make this cost low. In this paper, we have proposed various options which would contribute towards the GREEN savings. Energy Efficient Base Station (NodeB/ enodeb) A densely populated country such as India is currently connected by around 500,000 base stations. The new spectrum is getting auctioned in both 3G and 4G frequency bands and operators are expected to come up with more base stations offering next generation wireless services to provide higher peak throughput and lower latencies. On the other hand, almost none of the operators are showing any intent to phase out 2G base stations. As a result, the numbers of base stations are only going to increase in coming days. The following design considerations are made to reduce the energy consumption of the base station:

1) Reducing Power Consumption in case of Power Supply Failure The radio equipments such as RRHs (Remote Radio Heads) and Power Amplifier (PA) are usually inefficient devices in terms of power utilization. The typical RRH in the market have about 20% power efficiency and could attribute to about 80% of the power consumption of a base station (see Fig. 2). In case of grid electricity failure, the operator can choose to provide the electricity back up with battery/inverter. In case when the grid power failure happens and if the operators has multiple carriers, then the base station could detect such electricity failure and could reduce the number of carriers emitted on the site by switching off few of the radio equipments. This would decrease the power consumption of the site by about 40% and would increase the stand by time of the base station on the battery. If the operator has frequencies in multiple bands and when a grid failure is detected, the operator could choose either to switch off a cell on higher frequency band, or they could choose to switch off a radio equipment hosting less number of carriers. With such an approach, without compromising a lot on coverage and capacity, a considerable OPEX savings can be achieved. The Fig. 4 below shows the power saving/ reduced inverter / increased stand by time by switching off few of the RF assets on grid failure. power saving by switching off the rf assets 2500 In watts 2000 1500 1000 Power saving by switching off rf assets 4 3.5 3 2.5 2 1.5 1 0.5 0 number of 800w rectifier needed number of 800w rectifier needed Series 1 Fig 4: Reduced rectifier / Diesel back up need with switching of RF assets. 2) Renewable Energy Alternate Power Resource Usage To use the natural ways for managing the backup, for e.g. using Wind Energy OR using Solar Energy OR using alternate fossil fuel, and further use the output to charge the required backup.

3) Traffic Usage Monitoring Most of the sites have certain busy hours and certain lean traffic hours (Off Peak hours). For example, a base station covering a college campus could have busy hours in the morning and lean hours in the evening. The lean traffic hour could be supported by less hardware resources. There could be a mechanism in place to detect the Low Traffic Usage pattern.when there is no or very less traffic usage for a particular period of time, then those particular cells can be reconfigured/ hosted on another available hardware (which supports the particular frequency), and some of the hardware (/Radio resources) can be Shutdown. As soon as the traffic usage pattern resumes (which shall be monitored periodically), on cells / hardware boards, the switched off hardware boards/ Radio can be re-started again.upto 20% power saving can be achieved by switching off the additional hardware/ Radio resources of the base stations. 4) Use higher power Radio equipments to reduce the number of sites Most of the radio equipment in the market has a power efficiency of about 20%. If providing coverage is more important than having capacity (such as base station covering highways), then by choosing the RF equipment with higher power ratings would help in reducing the number of base stations needed to provide the coverage. This would directly help in reducing the CO2 emission and improve the TCO for the operator. 5) Use SDN - Multi techno Hardware to reduce number of sites Most of the operators across the globe have invested heavily in 3G technology and are yet to make some good returns on the investment. As 4G is already available in the market, it can co-exist with 3G system. A Software Defined Network (SDN) would eliminate the need for a separate installation, or managing a new hardware. Multi technology Radio Frequency (RF) hardware is available in the market which can support 2G, 3G, and 4G wireless technologies on the same hardware. 6) Use Carrier Aggregation instead of MIMO for 3G The MIMO provides a higher peak throughput and increases the spectral efficiency. However, MIMO also needs additional hardware. The Dual Carrier feature provides high peak throughput without the need of extra complex RF hardware. With the help of coordinated scheduling across the carriers, operators can achieve better load balancing and higher cell level throughput. 7) Use higher modulation for increased spectral efficiency Spectral efficiency can be improved with higher order modulation scheme. The peak data rate and cell capacity can be improved by going for higher order modulation. 8) Smart Antennas Smart antennas employ a set of radiating elements arranged in the form of an array, which enables a higher capacity in wireless networks, by effectively reducing multipath and co-channel interference. This is achieved by beam-forming / focus of the radiation in the desired direction, and by additionally adjusting this for the change in traffic pattern or signal condition. Energy Efficient Controller To cater to the GREEN Wireless Solution, there can be optimizations also at the Radio Network Controller, the node which is a controlling element in the RAN which connects with the base stations. For such cases, there are ways to improve the running cost using the following options: 1) Power Failure Detection In case when the power failure happens and operators have multiple carriers, they can optimize using a mechanism to detect this power failure, and accordingly, stop the operating hardware for a particular carrier (if possible higher frequency carrier), as this would reduce the load by 1/2 and reduce the overall power usage at the Base Stations

2) Cell Load Sharing When operators have more carriers (>1) supported in same coverage area, then, under low traffic situations, the operators can switch off some of the cells and provide the services using other available cells in the same coverage. 3) Traffic Monitoring When for a particular period of time, there is no traffic on a particular cell, then there can be a mechanism in place for shutting down the cells resources in such a way that the users on the cell which is going down are not impacted and those users can silently be migrated to another available coverage cell. As soon as the traffic pattern increases, the suspended cell can be put in service. 4) Coverage Optimizations Certain operators have more than one carrier operating under two different frequency bands. In such cases, the operators who are using Cell Load Sharing technique (captured in point 2 above) shall switch down the cell with higher frequency, as higher the frequency the coverage, lower the penetration. So with the lower frequency cell available, the coverage and penetration would not be impacted. 5) Better Planning Cell size can be planned properly so that the unnecessary power dissipation is reduced, and we use the RF & digital resources efficiently for transmission on the Air Interface. 6) Sleep Mode Wireless Systems can be designed to support sleep mode, so when there is no transmission happening, it can fall back into sleep mode, reduce unnecessary power dissipation, and reduce the interference. 7) Self Organized Networks (SON) With the continuous growth of mobile users, increase in data traffic, and higher requirement of managing the capacity, network operators need to deploy large number of base stations with a reduced coverage. This large number of base stations will also add on the energy consumption of the wireless networks. By using the Self Organized functionality, the networks can remove manual configuration during deployment or during load conditions. The networks can also adapt dynamically to manage the capacity, by adding/ removing the cells dynamically. This is done by adapting the need for the transmission power/ changing the neighboring cell information, using MIMO configuration. 8) Use RAN sharing As described in Fig.1, RAN consumes most of the power in the wireless network. Two or more operators can go for the RAN sharing solution, where the RAN is owned and managed by one operator, but shared with other operator on fixed cost or pay-as- you-use basis. 9) Swith Off some RAT system during low traffic The existing 3G system can give a throughput as high as 42 Mbps in DL and 11Mbps in Uplink. The operator having the 4G overlay on the 3G network, can choose to switch off 4G network, while under low traffic hour. Such a decision can be automated by the centralized multi technology management system or handled automously through a SON based approach Alternatively, operator with good 4G coverage could use 3G network for capacity purpose and could switch off the 3G network doring low traffic hours.

RAN Optimizations Transport / Backhaul / Cabinet Optimizations Transport/ Backhaul does the transporting of traffic between distributed sites (typically access points) and other nodes connecting nodes and cabinets are designed based on the nodal requirement. They also contribute significantly towards the power consumption. For managing theses nodes/ interfaces we have proposed various options which would contribute towards the GREEN savings. The RAN part accounts for more than 50% of the total power consumption in the Wireless Systems, so operators should try to keep on optimizing and make this cost low. In this paper,we have proposed various options which would contribute towards the GREEN savings. BBU Hotelling In the BBU Hotelling solution, the Base Station (the controlling part) is separated from the RADIO Remote Head (RRH) equipment. All the Base Stations are configured/ kept in a single place Hotels. As all the BBUs are placed in single place, the day to day need for the site visit is reduced. Additionally the overall cost of running & cooling the systems is optimized and thus helps in reducing the energy requirement. Going forward, most of the operators target to plan the system using this architecture for improving energy efficiency and reducing TCO. Reduce the Heating cost/ Auto Cooling Telecommunication hardware needs to be kept running at all times. Continuous running heats up the systems, for which there are mechanisms required to keep the systems cool (under control) to get expected results all the time. This can be achieved using the Fan systems cooling or using Auto Cooling. Using the Fan requires additional cost, so as much as possible, auto cooling options can be explored. Although in some cases a fan is required, where the hardware / cabinets are not readily accessible. Choosing the ICs, FPGA with higher operating temperature would reduce the need for cooling. Instead of keeping the fixed speed fan for cooling, if the fan speed could be adjusted on a need basis, then some energy saving can be achieved. CPU throttling / Dynamic Frequency Scaling Any device when operated at a higher clock rate, requires more power. Hence, reducing the clock rate of the microprocessor through power management based on the utilization / usage will reduce the energy consumption as well as power dissipation. So this will (1) conserve the Power & (2) reduce the power required due to less heat generated by the CPU. A majority of the processors currently available provide this facility, during the off-peak hours. The operating frequency of the CPU can be reduced to bring down the power consumption. Another option is to use only single core out of a multi core processor dynamically, based on the load conditions. If required, frequency scaling shall be done on core basis. Fig.5 shows the frequency scaling behavior against the utilization [5].

Saving per board per hour may be low (e.g. about ~5W) when using the frequency scaling / Core shut down. Due to the economy of scale (i.e. number of cards in the network), this becomes critical and the overall saving becomes considerably high. Performance Max Frequency (ghz) available frequency Min Ondemand Power save 20 80 100 Utilization(%) Fig 5: Dynamic Frequency Scaling VS Utilization Power Transmission For the connection from the Base Station (i.e. BBU) towards the Radio Equipments (RRH), it would be better to use AC instead of using DC, and at the Radio Equipment, this can be converted back into DC and fed back into the equipment. Most of the Radio uses -48V power input. When the operator uses the BBU hotelling kind of solution, the distance between the RRH and base station would increase. Transferring low voltage DC power leads to higher power dissipation. In such cases, it would be better to use AC instead of using DC to RRH, and at the Radio Equipment, this can be converted into DC and fed back. Use tower mounted Radio to eliminate feeder loss The traditional macro base station solution had a long RF feeder cable running from the TransReciever (TxRx) module to the antenna, and accounts for about 1.3-1.5db Tx Path loss. In other words, about 30% of the transmitted power is lost before reaching the antenna. Using the distributed base station solution with RRHs could help in getting the radio module close to the antenna. Thus it eliminates power loss and help in increasing the coverage. Conclusion In this paper, we have discussed various proposals which provide GREEN solutions for the wireless networks. Wireless operators, access networks, and base stations become important and shall drive technology to energy efficient directions. Improving energy efficiency and reducing the overall power can be provided as a generic solution. All the wireless electronic equipments involved directly or indirectly with the end user and the access network can optimize thermal processes, efficient network planning, and base station design, to reduce the overall footprint.

References 1) Recommendations on Approach towards Green Telecommunications (Telecom Redulatory Authority of India 12 April, 2011) 2) Green Mobile Power & Community Power Project (GSMA Resources : http://www.gsma.com/mobilefordevelopment) 3) 3GPP Green activities / Energy Saving (CP 120366) 4) 3GPP 25.927 Solutions for energy saving within UTRA Node B 5) Manish Bansal, Deepak Pushpan, Hari Medha, Saket Jain, Sumit Jain, CPU Frequency Scaling by Utilization, Technical Report IIITB, April-2012 Author Info Mohit Chouksey Mohit Chouksey joined HCL over 1 year ago in the Engineering and R&D (ERS) & has overall 15+ years of experience. He has extensive wireless domain expertise in Architecture & System Engineering. Nithin Kumar Nithin Kumar joined HCL over 1 year ago in the Engineering and R&D (ERS) & has overall 10+ years of experience. He has extensive wireless domain expertise in Architecture & System Engineering. Hello there! I am an Ideapreneur. I believe that sustainable business outcomes are driven by relationships nurtured through values like trust, transparency and flexibility. I respect the contract, but believe in going beyond through collaboration, applied innovation and new generation partnership models that put your interest above everything else. Right now 110,000 Ideapreneurs are in a Relationship Beyond the Contract with 500 customers in 31 countries. How can I help you? TM