Abstract
Power consumption became a crucial problem in the development of mobile devices, especially those that are communication intensive. In these devices, it is imperative to reduce the power consumption devoted to maintaining a communication link during data transmission/reception. The application of dynamic power management methodologies has contributed to the reduction of power consumption in general purpose computer systems. However, to further reduce power consumption in communication intensive real-time embedded devices we have to consider the state of the computation and external events in addition to power management policies. In this paper we propose a model of an Extended Power State Machine (EPSM), where we adapt a Power State Machine to include the state of an embedded program in the power state machine formulation. This EPSM model is used to adapt the Quality of Service (QoS) in communication intensive devices to ensure low power consumption. In such development, a middleware layer fits in the system’s architecture, being responsible for intercepting the data communication and implementing the EPSM. Also, a software tool was developed, allowing the Middleware Code to be generated based on the State Machine. A case study demonstrates the application of the proposed model to a real situation.
I. INTRODUCTION
Due to the evolution in the component miniaturization process and the development of high-speed wireless network technologies, there was an increase in use of mobile devices in the last few years. The development of high-speed wireless network technologies enabled an increased demand for mobility and changed the focus of computing systems. These systems are now communication intensive, which means communication tasks are the system’s main goal and they must be executed under certain constraints. Sensor networks and wearable computer applications are examples of this shift in paradigm. The increasing use of mobile devices lead to the development of several applications and to an increase in these systems functionalities. However, the power consumption problem was accentuated once these devices are battery operated and we know that batteries are not inexhaustive. Treating the power consumption problem on PCs is quite different, once we suppose we have a continuous power supply. On laptops, otherwise, we do not have a continuous power supply, but their size allow the use of bigger batteries, providing consequently more autonomy. The power consumption problem arises since it is desirable to have mobile devices as small as possible within longer periods of time, and battery’s capacity grow with battery’s size. In this work we will consider, more specifically, real-time embedded systems used in a communication intensive environment, which have some peculiarities that may be considered when we are developing new applications. For real-time constraints we mean that a real-time system must satisfy explicit (bounded) response-time constraints or its correctness may be compromised, risking severe consequences, including failure [1]. In these systems, the response time is as important as the correctness of the outputs. A real-time system does not have necessarily to be fast; it must simply produce correct responses within a definite time limit. A real-time embedded system usually monitors on the environment where the embedded system is installed, and if it does not respond in time to a request, the result can be disastrous. Examples of real-time embedded systems are aircraft engine control systems, nuclear monitoring systems and medical monitoring equipment. The need for real-time responsiveness associated with the communication intensiveness introduce extra constraints to a power management policy. The embedded systems design considers the systems characteristics and restrictions that are fundamental for an efficient system function. As a result, low power design of communication intensive real-time embedded systems must consider the environment and application constraints to optimize the system’s design [2], such as real-time responsiveness and intensive execution of communication tasks. The interaction between the system and the environment may be represented by external events, which must be considered when reducing the power consumption. It is also important to consider the state of computation when the system turns on/off components to reduce power. The state of the computation in each period of time represents the state of the application and its restrictions in an instant of time, which can have a direct influence on the decisions made by a power manager. Although energy minimization of embedded and mobile computing is of great importance, energy consumption must be carefully balanced against the need for real-time respon- siveness. And more important than the power, delay, or even energy is the relation between energy and delay (E x d), once it represents the energy that is spent executing a task. A power manager task can act in a system by turning off components or adjusting the QoS parameters according to the state of the computation, without harming its timeliness. As an example we can mention an application for remote cardiac monitoring system. In this case it is very important not to interrupt the monitoring function when an abnormal condition is detected, even if the battery power level is critically low. In this work we proposed a dynamic power management model for a communication intensive real-time embedded system based on systems’ quality of service (QoS). This model considers the application’s constraints and the environment for the implementation of a dynamic power manager. QoS is defined as a function of the battery power level and the need for system responsiveness (here captured by the computation state) at an instant of time. We propose a model of an Extended Power State Machine (EPSM), where we adapt a Power State Machine to include the state of an embedded program in the power state machine formulation. This EPSM model is used to adapt the Quality of Service (QoS) in communication intensive devices to ensure low power consumption. The model implementation was done through the development of a middleware layer that incorporates a power state machine and executes as a thread independent of the application. The middleware layer hides the difficulties deriving from the combination of diverse applications implementing the same power management model, once it may control the hardware accesses from all applications. We have also developed a graphical tool where the Extended Power State Machine is defined and the code for the middleware layer is generated. A real application where the model was applied is also presented in this paper. This paper is organized as follows. In Section II we talk briefly about our motivating example, a wearable computer based physiological signal monitoring, which is presented in detail as a case study in Section IV. In Section III we first define some basic concepts and present a new model proposition based on Power State Machines, and show the architecture of an implementation through a middleware layer supporting the EPSM, presenting also a tool created to generate the middleware code for both client and server sides of the application. In Section V we present a case study of power consumption reduction in a wearable computer
MODEL IMPLEMENTATION THROUGH A MIDDLEWARE LAYER
Considering a communication intensive application that consists of a server and a client exchanging data through a wireless link, a middleware layer was developed implementing a power management policy while controlling the applications access to hardware and software. The middleware intercepts the data exchanged between the application and the communication interface and adapts it according to a power management policy. The middleware layer is incorporated in both client and server, once the data transmission is processed according to the power management policy. This approach is interesting since the presence of a power manager in the system is transparent to the user, although it demands an adaptation from the application side. The development of a middleware layer simplifies the application’s implementation, since it does not have to take care of the power management tasks nor worry about other applications that could be accessing a hardware that it is trying to turn off. Figure 3 shows how a middleware layer fits in the system architecture. The dashed line represents the use of a wireless communication interface in the communication process between the mobile unit and a fixed unit. The use of a mobile unit and a fixed unit is just an example and might be extended to other combinations of mobile and fixed units. Our choice for a mobile and a fixed unit was based on our motivation example and it expresses clearly the power management problem present in mobile units of all kinds. The insertion of a middleware layer in the system’s architecture As just explained, the middleware layer is responsible for intercepting the data communication and implementing the Extended Power State Machine. Although all of the control of the state machine remains in the middleware, the application is responsible for defining the EPSM structure to the middleware according to the tasks executed at each application condition or state, including the states and transitions. A. System Architecture The application defines the states of the state machine and their respective QoS attributes and this state machine is the core of the middleware. The application defines also the transitions, indicating, for each transition, a condition test function and the next state it takes in the state machine if this condition is satisfied. Thus, the middleware is responsible for controlling the transitions between states and sending/receiving data to/from the communication interface according to some state QoS. Some concepts were defined for a better understanding of the system’s architecture and they are just presented
• Data buffer: a data buffer is a shared data structure used for the data exchange process. The application writes data to data buffer while the middleware reads data from the data buffer. This read/write access is controlled through the use of mutexes.
• Data block: a data block consists of a set of data grouped according to its transmission rate and maximum transmission time characteristics, on which will be applied the same procedures for data transmission and storage. For each data block we have one data buffer associated, and all the data for that data block is transmitted using this data buffer. Also associated with a data block we have data map functions, which are application-defined functions that provide an adjusting mechanism for the transmitting rates. For example, an application may be filling the data buffer in a higher rate than the middleware is capable to send data to the communication interface. In this case, an adjustment is necessary, such as a medium, an intermediary value, or other, according to the application needs. The data map function is responsible for applying this adjustment. These functions could also do some kind of backup, like saving not transmitted data in a file for posterior analysis.
• State: a state of the state machine can be defined as a group of data blocks with different transmission times and transmission rates that indicate how the data in their respective data buffers must be sent to the communication interface. We have also associated to a state a list of transitions to be tested and executed.
• Transition: a transition consists of a pointer to the next state and a transition test function, which is defined by the application as a condition test function
CONCLUSIONS
In this work we proposed a model to achieve low power consumption in real-time embedded systems based on dynamic power management, where the state of the computation is as important as the battery capacity. The formulation was called an Extended Power State Machine (EPSM), where the states of an EPSM depend on the embedded program’s state, and the transitions depended on external events. Power consumption was reduced based on adapting the quality of service for the variables that were communicated between a mobile and a fixed unit (monitoring station). Each state of the system required a different QoS. A case study was presented for a medical wearable device that considers the patient’s state when deciding upon the quality of service and responsiveness for the sensors’ data transmitted over a wireless network. We showed that significant power reduction in communication was achieved when this technique was applied. The EPSM was designed as a multi-thread application comprising the medical application and a middleware layer, which implements the EPSM. The solution adopted on designing a middleware layer facilitates the implementation of further applications.
Iot based engineering projects
iot based engineering projects
iot based projects for engineering students
iot based major projects
iot based major projects for ece
iot based projects for mechanical engineering
iot based final year engineering projects
iot based projects for final year engineering students
iot based final year projects for electronics engineering
iot based project
iot based best projects
iot based electrical engineering projects
Thank you for contacting Shield Technologies! Please let us know how we can help you.
For more information please do visit mifratech center.
Note : Find the best solution for electronics components and technical projects ideas
keep in touch with our social media links as mentioned below
Mifratech websites : https://www.mifratech.com/public/
Mifratech facebook : https://www.facebook.com/mifratech.lab
mifratech instagram : https://www.instagram.com/mifratech/
mifratech twitter account : https://twitter.com/mifratech