Published Aug. 14, 2019, 1:48 p.m. by Moderator


This paper discusses power-aware OS/algorithm for handheld devices. An operating system commonly referred to as “OS” is a central program that operates the computer. It is the main program which ensures that all the other programs and/or applications operate. An operating system can perform many functions from basic ones like distinguishing input from the keyboard and mouse to display an output. It could also be connected to other additional devices like a printer. In production servers which use powerful computers that perform billions of calculations every second, the operating system keeps the different competing programs working harmoniously. It also protects the server from unauthorized access. (Beal, 2019).

Operating systems are categorized into four main categories; real-time operating systems, single-user/single task operating systems, single-user/multitasking operating systems, and multi-user operating systems. Real-time operating systems operate like a ‘sealed-box’, there is very limited user interaction and the computer is tasked with performing periodic tasks that have to happen at an exact time for a defined period every day. These operating systems are used to operate industrial plants, heavy machinery or scientific equipment (Curtin & Dave, 2019).

Single-user or single tasks operating systems can only be used by a single user at any particular time. A good example is the Palm operating system that used with the Palm handheld computers. Single-user/multitasking operating systems are the most common of operating systems. Good examples are Microsoft Windows and Apple’s Mac OS. They allow a single user to be able to run multiple programs at single time depending on the computer’s hardware. A multi user operating systems allow many users to access the computer simultaneously. The operating system balances the needs of each user depending on the task involved (Curtin & Dave, 2019).

Critical Analysis

The three papers chosen for review are “Exploiting Energy Efficient Emotion-Aware Mobile Computing” by Yuyang Peng et al (2017) which delves into emotional aware computing that focuses on the emotional state of the user.  The second paper is “Using the big data generated by the smart home to improve energy efficiency management” by Rodriguez Fernandez et al (2016) where they propose using machine learning algorithms that encourage the user to be more energy efficient. The final paper of the review is “Energy Optimization for Mobile Device Power Consumption: A Survey and a Unified View of Modelling for a Comprehensive Network Simulation” by Benkhelifa et al (2016) which compares energy consumption levels across different devices with an emphasis on mobile devices.

In their paper “Exploiting Energy Efficient Emotion-Aware Mobile Computing”, Yuyang Peng et al (2017 can be described as creating a new sub-culture of mobile computing which focuses on collecting huge amounts of data from users, for the algorithm to make accurate recommendations to users based on their emotions. Implementing such a model requires hardware and/or software that consumes a lot of power. Yuyang et al proposes two approaches; the first one focuses on the user by encouraging energy efficiency and the second one focuses on remote data centers that will focus on using renewable energy (Yuyang et al, 2017).

The paper describes that to achieve emotional aware mobile computing a lot of data needs to be collected. Data collection and data processing are energy intensive. To achieve energy efficiency, the actual data processing is done on a remote cloud server not the user’s mobile phone. 5G network has enabled smartphone and IPhone users to be able to access more personalized mobile services. Voice recognition and home surveillance systems are now available on handheld devices (Yuyang et al, 2017).

Yuyang et al (2017) classify the data to be collected into two categories; physiological and physical measurements. Physiological measurements are emotional states that are measured through brain activity, heart activity, blood pressure, breathing, muscle activity and skin responses. Physical measurements include facial expressions, voice intonations, eye activity and surrounding environment which can be measured using mobile devices, sensors and cameras. Physiological measurements require more sophisticated equipment’s such as MRI’s and electromagnetic devices (Yuyang et al, 2017). They further reiterate that energy efficiency is a very important consideration for the future of emotion aware mobile computing.

This is a challenge as the machine learning algorithms in their model require a lot of data to be able to make accurate predictions and the tradeoff between functionality and power consumption is an obstacle to achieving this. It is very important to achieve a balance between power consumption and functionality. In the renewable approach, the researchers propose that machine learning models that predict weather can be used to maximize energy generation. Wind turbines energy can be harnessed using wind speed and power and solar energy using temperature. 5G networks offer faster speeds and better coverage with very little downtime.  Some of the proposed ideas to make 5G networks energy efficient include millimeter wave that extends the bandwidth of 5G increasing its coverage, ultra-dense cells to improve energy efficiency and MIMO technology that is used to improve the spectrum efficiency in 5G mobile communication (Yuyang et al, 2017).

The second paper in the review is ““Using the big data generated by the smart home to improve energy efficiency management”, by Maria Rodrigues Fernandez et al (2016). In this paper, Maria et al discuss the use of handheld mobile devices or Internet of Things (IOT) to create a smart home which records electricity consumption and temperature. The data collected is used in machine learning algorithms to understand the user’s consumption habits after which the computer will make recommendations and predict future user behavior. This is the prototype of a ‘smart grid’ system that monitors power consumption over a national grid.  The first option to achieve a ‘smart grid’ system; the algorithm is able to predict user behavior and during off peak hours, the system will shift the customers’ allocated power load to a customer who has a higher demand. The second option is to encourage customers to be energy efficient by their own volition (Rodrigues et al, 2016).

Maria Rodrigues Fernandez et al (2016) further explain that ‘Smart Metering’ is an inexpensive way of engaging the end consumers (Rodrigues et al, 2016). The power consumption monitoring device records information about the user’s consumption. The user will then use the collected information to find better ways to manage their power efficiently. It assists the user to efficiently manage power consumption in all their devices by turning off idle or unused devices. Sensor metering where sensors are placed strategically in different rooms to monitor temperature & humidity enhances the scope and quality of the data to be used in the machine learning algorithm (Rodrigues et al, 2016). This is like a miniature ‘smart grid’ system designed for domestic use. One of the benefits of the ‘home smart grid system’ is the user can detect illegal power connections based on their consumption.  With increased data the algorithm will be able to make predictions and recommendations. The power companies can use the predictions to improve efficiency and the domestic user can use the information to reduce energy consumption and save costs (Rodrigues et al, 2016).

The final paper for review is “Energy Optimization for Mobile Device Power Consumption: A Survey and a Unified View of Modelling for a Comprehensive Network Simulation” by Benkhelifa et al (2016) gives a detailed analysis of energy consumption rates on different devices across different networks and then proposes a ‘sweet spot’ where energy consumption is most efficient. Benkelifa et al (2016) write that to be able to create an energy efficient operating system or algorithm. It has to take into account the consumption of the different components which enable the device to work. Ben Khelifa et al (2016) explain that a healthy balance must be achieved to ensure that all the different components fit and work together to achieve energy efficiency. Ben Khelifa et al (2016) write that Wireless Sensor Networks prefer energy efficiency across all the devices as its performance depends on the processing speed of the connecting device. Even the protocols or connecting settings are geared at achieving energy efficiency. Energy inefficiency in one device compromises energy efficiency in other connecting devices (BenKhelifa et al, 2016).

The paper explains that in terms of devices; laptops are geared more at memory intensive tasks and so their primary goal is functionality rather than energy efficiency while smartphones are created for mobility rather than functionality (BenKhelifa et al, 2016). Energy efficiency is key requirement in achieving mobility and longer battery life is the only solution. However computers and mobile devices have hardware and software that changes power consumption based on requirements. Cloud computing servers require huge amounts of power and energy efficiency can only achieved through renewable sources as they can only perform with direct power from a grid. It is important that energy efficiency should be achieved in mobile phone technology as they are the most widely used (BenKhelifa et al, 2016). To achieve energy efficiency in mobile phones there are three areas to consider; network, hardware and software. Hardware includes processors, storage devices, communication devices, and sensors while software includes the operating systems and user installed applications (BenKhelifa et al, 2016).



A power aware operating system or algorithm has the ability to manage power consumption efficiently. By using less power, the battery life of a device is prolonged. The challenge of insufficient battery life is most common in hand held devices where the user does not need to plug the device to use it. Examples of handheld devices that do not require direct charging include smartphones, IPhones, tablets and IPads. Increased network coverage of 3G, 4G and 5G networks has increased the accessibility of smartphones. Smartphones are now the most widely used handheld devices. They are small in size and light weight and their portability makes them popular with users. Networks like 4G and 5G have increased the performance in smartphones. Android and IOS applications enable a smartphone user to perform tasks that were exclusively done on computers. This increased performance requires more power. Finding the perfect balance between functionality and power management is an area of great interest to many researchers. Smartphone batteries have a limited shelf life and it is important to prolong the battery life. In terms of functionality, key considerations include backlight, number of processors, memory size, and number and type of applications on the device.


Benkhelifa, E., Welsh, T., Tawalbeh, L., Jararweh, Y., & Basalamah, A. (2016). Energy optimisation for mobile device power consumption: A survey and a unified view of modelling for a comprehensive network simulation. Mobile Networks and Applications, 21(4), 575-588. doi:

Curtin Franklin & Dave Coustan. (2019). Types of Operating Systems | HowStuffWorks. Retrieved March 18, 2019, from

Rodríguez Fernández, M., Cortés García, A., González Alonso, I., & Zalama Casanova, E. (2016). Using the big data generated by the smart home to improve energy efficiency management. Energy Efficiency, 9(1), 249-260. doi:

Peng, Y., Peng, L., Zhou, P., Yang, J., Mizanur, S. M., & Almogren, A. (2017). Exploiting energy efficient emotion-aware mobile computing. Mobile Networks and Applications, 22(6), 1192-1203. doi:

Vangie Beal. (2019).What is an Operating System - OS? Webopedia Definition. Retrieved March 18, 2019, from

 ⋅  0 comments have been posted.

Post your comment

Required for comment verification