Mobile consumption. This research paper will present the

Mobile Processor Energy Consumption and UsageUP746831(Andrew Mitchell) UP781408( Nattasan Chantaramanee)Abstract—The purpose of this research paper is to look at the energy and power consumption by mobile processors. How the size of the processors affects energy consumption levels. Study how mobile processors have developed over the years, the growth of mobile functionalities and the impacts on the end users. We present a comparison of different brands of the mobile processors and analysis breakdown of processors usage and energy and consumption. In recent years, hardware architects have found multicore processors the way to achieve better performance with lower power consumption. We will identify how hardware developers have managed to take advantages of current technologies and construct mobile processors that enable mobile devices to perform high performance with relatively low energy and power consumption. This research paper  will present the results of energy and power consumption experiment conducted on available smartphones. The test cases includes some of the daily tasks/ functionalities on the smartphone. The results obtained  were used to draw the conclusion.——————————   ?   ——————————1 IntroductionHandheld devices like cellular phones and PDA (personal digital assistance) have been growing in terms technology and demand since the first one was developed back in 1973 by a company called Motorola. The first handheld phone was, in comparison to what is considered large today, is huge as well heavy; weighing in over 1kg and having 30 minutes of talk-time, six hours standby, and could store 30 phones numbers. It also cost £2639. These features seem very minimal compared to the technology that is being used today. For example, the Iphone X size fits into a hand, weighs 174 grams, 21 hours of talk time, over 48 hours standby time and with near unlimited amount of features and customisation. The reason was due to technology at the time but now there is a new problem that has popped up. It is how mobile processors are getting smaller and more energy efficient but in reality this is not the case. This research paper will look at the development of mobile processors and how they are linked to energy consumption and efficiency with current and future mobile devices as well as review the methods that are being used in the creation of safer and energy efficient mobiles2 RELATED WORK OR LITERATURE REVIEW2.1 Review Stage 1Power and energy consumption of mobile processorsToday’s device processors are the result of the generational microarchitectural enhancements, memory hierarchy growth, clock frequency increases and multicore scaling. Mobile processors are part of the innovation, having embraced these architecture features in pursuit of top performance. For the purpose of this research paper, we measure the impact on performance, power and energy on some of the most used processors in the mobile phones released from 2010 to 2017.An import aspect of conducting this study is selecting the right “sample” to study. Our research paper focuses on the mobile processors released between 2010 and 2017.Fig 1: Mobile processors overview released from 2010Fig 1: Mobile processors overview released from 2010Figure 1 is an overview of the major processor’s families that have been released since 2010. The specifications have evolved significantly over time to prove mobile phone capabilities and functionalities to satisfy end users. However, such performance enhancements have resulted in excessively high power consumption.In early days, mobile phones come with single-core processors where they did every task single-handedly. Until several years ago, Fig 1, where manufacturers have taken advantage of modern technology and introduced multi-core processors. Multi-core processors are designed to divide phone’s task among multi-cores with the purpose of completing the task faster.  Although, hardware advancement has improved; but now the growth of application functionalities demand even more performance, the issues of heat dissipation still exist in some industry as manufacturers could not make processors any faster in a tiny space and power drainage factor in consideration. Furthermore, the fact that mobile devices operate from battery power which makes it a major factor in power consumption. Fig 2: Battery capacities of mobile phones.With phones being released every year, the latest generation seems to come with a better capacity and increase in battery life; however, the bigger the battery there is the more energy required. To satisfy the end users, mobile phone manufacturers design for the battery to last as long as possible to cope with device functionalities and capabilities. Nonetheless,  battery capacity is restricted due to the constraints on weight and size of the device itself. This suggests that energy efficiency of the devices is seriously taken into account by the manufacturers in order to meet users’ expectations.Fig 3: Device Minimum Screen-on powerFigure 2 shows the power consumption on some of Android devices has been released in recent years. The table represents important power value that is unavoidable constant drain of energy whenever the actively being used. There is a noticeable trend which shows power consumption slightly reduced as the development of the  mobile processors. For example, Galaxy Note 4 (Exynos 54330) which has the screen on power consumption at 452mW; whereas Galaxy S6 (Exynos 7420) equipped with an upgraded processor that was released a year after has a screen on power consumption at 358mW. There is evidence that the progress is being made in terms of technology advancement while making sure that power consumption is taken into account and keep low as possible. However, if you take a look at the Galaxy S5 and Galaxy S5 LTE-A, the Galaxy S5 LTE-A was released after the Galaxy S5 and has more powerful processor than the S5 and it consumes more power; which indicates  that recently established technology does  not necessarily mean it provides the solution to high performance with low power.2.2 Review Stage 2The mobile industry is experiencing huge growth worldwide with around 7 billion subscribers and an ever-growing demand for data traffic. With this huge use and growth of mobile usage, one would assume that the network and mobile device itself would be efficient but this is not the case. Both base stations and smartphones regularly waste 70% of energy consumed as heat. This is due to outdated technology that was developed during the 1930s. The European Union has funded a project called GreenNets that has been developed specifically for the solution of mobile networks to cutting energy costs and reduce Co2 emissions. They do this by creating an automated network management software tool that continually adapts network activity during quiet periods, saving energy. They also found out ways to evaluate mobile network efficiency.Figure 4: Daily TrafficFor example the graph above shows the daily traffic of mobile devices within specific zones as well as a huge decrease in traffic during the weekends that can be used to evaluate periods of times to reduce power consumption of the mobile network therefore reducing Co2 emissions and running cost to the company without compromising the quality of service for the end users. To implement this it would simply be a functional extension of the mobile phone providers already existing which would lead to no investment in new hardware or technology. If working this technology could and would improve energy efficiency substantially but there are too many variables that can affect the network management tools monitoring. For example factoring in the size of the network the tool would have to monitor would be constantly shifting due to ever growing consumption as well as unforeseen interruptions ‘65% of communications interruptions are caused by power supply failures and 85% of them are discovered after more than 12 hours, due largely to customer complaints.’ Increasing energy efficiency in mobile networks. (2014). Retrieved from 1. These interruptions could be mistaken as traffic decrease and could lead to the wrong power output causing even more disruptions in the network. Another project that is more realistic and currently in use by Nokia is ETA devices. Eta devices have developed a power management technology called ETAdvanced that uses an amplifier to significantly reduce heat waste by using a system like an automated gearbox, constantly tweaking energy usage to provide only enough power for a radio signal. It is said to have extended battery life in smartphones by 50% and as a result 80% reduction in heat waste as well as improving mobile base stations to 70%, which is 30% to 60% more efficient than current base stations as well as savings of up to 50% to macro base stations for a direct impact on operating expenses. Summed up this method of approach to improving mobile energy is based around the modifications of mobile base stations by installing an amplifier which in turn improves mobile devices that can use the newly acquired radio signal from the amplified base tower. As well as allowing for the amplified signal to improve energy efficiency for connecting mobile devices it also improves the base station itself, this is done by powering small cells (e.g. pico and femto cells) with ethernet cable alone as well as since the smaller cells need less room that translates as smaller cabinets needed causing cheaper rental costs and more location options. They can also be solar powered base stations for developing countries to further reduce running cost. The problem that occurs when installing new tech is the new tech itself, to make the energy more efficient requires scaling it back at the base stations by adding smaller energy cells thus picocells femtocells. With the decrease in size comes a decrease in numbers of supportable subscribers as well as the smaller output in power will lead to limited area coverage, as shown by the diagram below.Figure 5: Small Cell BasicAnother problem with setting up the small cell networks is establishing the appropriate backhaul. This is very time consuming as finding the appropriate location to set the low-level access as well as uncertain deployment time for small cells due to site acquisition and zoning. A more physical and practical improvement that has been a major development in mobile phones is Multi-core processing, as stated above. When it comes to energy efficiency the multi-core has multiple cores in which tasks on the mobile phone can distribute on different cores but this comes with problems, like the ones stated already, such as heat dissipation due to the small space for the new cores. Even though there are benefits to the multi-core such as improved speeds and performance but with this there are disadvantages such as due to higher power there is less battery life in the mobile device and more heat energy waste. This ‘improvement’ is not really required for the average phone user that just texts, calls and uses it for social media applications. A more powerful phone is to ‘primarily appeal to tech industry insiders and mobile device enthusiasts. “It’s like the car hobby mentality of the guy who wants a dual overhead cam engine. But I’ll argue that with a smartphone, users can’t define what dual-core is.”‘ Do mobile devices really need multi-core chips?. (2012). Retrieved from 2. There are exceptions for example users who use their mobile device for work and need the improved power.3 Proposed solution/mode/approachFor the solution to energy efficiency, the focus will be on the mobile battery and the running time with different applications. If applications can be tested and monitored for their energy consumption then optimisation of the applications can be applied to improve the battery life. To measure the power usage the mobile phone would run as standard but be using applications such as Voice call, SMS, Camera, Youtube etc…Whilst running the applications it will be monitored using energy monitoring software. Key: where Pn denotes the power consumption of n coresPc(j) represents the power dissipation of a core j.Add = Additive powerAct = Actual powerPmc = The power consumption of chip-level mandatory componentsTo monitor the power consumption of individual cores j(Pc(j)) is first measured when it performs computations and the others remain idle. To compute power of n cores (Pn),Pc(j) is added n times as in Equation. Then present such a power as additive power consumption (Add) of n cores.Then compute power consumption by Pmc=Add?Act.After the energy usage has been calculated then adding energy efficient mechanisms can reduce the power consumption of mobile processors by dynamically scaling its voltage and frequency as well as by turning off unused components through a set of CPU-Bound and I/O bound sets of code. 4 Experiment/Result/SimulationHere we conducted the research on a smartphone device to show the energy consumption for some of the most used services/functionalities. It is necessary to use hardware-based energy consumption measurements, we opted to use a multimeter power monitor to measure the range of mobile battery and energy consumption while the device performs its functionalities.Three smartphones used in this experiment are:iPhone 7Samsung Galaxy 7 Fig 6: Power monitor device used for energy measurements.Fig 7:  Software tool for power monitor device.Test CasesTo obtain the measurements, test cases are produced for performing different functionalities. Description of test cases are as follows:1. Voice call – Ensuring the necessary ports are connected properly, then initiate the voice call. The measurement of energy consumption will be recorded within 60 seconds of the test period. 2. SMS – Short Message Service (SMS) was chosen for the second scenario to analyse the effect on energy consumption. The sample text was random and consist of 300 characters. So, the test runs when after “Send” button is pressed, and stops when the text is sent.3. Internet Tasks –  As browsing Internet and social media has become part of our daily tasks when using smartphones, this was chosen to be in test cases. In this case, we use Facebook and Googling in general to conduct the test. The applications will be performed for 60 seconds, while energy consumption is monitored in the background until the application is closed.Test resultIn this experiment, the units we use to represent the energy consumption are Joules(J). Test results are as follows:Voice call iPhone 7Galaxy S 7Energy consumption45.22 J41.24 J Fig 8:  Energy consumption result  by voice call SMSiPhone7Galaxy S7Energy consumption2.01 J1.54 JFig 9: Energy consumption result by SMSInternet taksiPhone7(browsing Internet, Facebook)Galaxy S7((browsing Internet, Facebook)Energy consumption50.54 J, 62.44J 49.01 J, 58.23 JFig 10:  Energy consumption result by Internet Task5   ConclusionOverall the research shows the current state of mobile processors with them being much improved since the start of mobile technology but nowhere near perfect. Single-core and Multi-core processing has improved the power of mobile devices but has raised the question about the energy efficiency of mobiles. With the increased power comes increase energy waste through heat dissipation. Work in energy efficiency through hardware and/or software is being developed which can help increase energy efficiency to some extent but there are physical limitations with the mobile base stations. With an increase in energy efficiency not only will the mobile be cheaper to run and manufacture also electromagnetic radiation will be lowered causing less damage to the environment as well as more efficient batteries translating to less time charging which reduces the energy required from the grid.8 ReferencesIncreasing energy efficiency in mobile networks. (2014). Retrieved from mobile devices really need multi-core chips?. (2012). Retrieved from 1: Figure 2: 3: 4: 5: