Citation Context

...come increasingly evident in both mobile and enterprise environments. Limitations in battery capacities and demands for longer device lifetimes have motivated research for mobile and embedded systems =-=[12, 34, 37, 38]-=-. For server systems, online management is important for two significant reasons. First, there are power delivery and cooling limitations in datacenter environments that result from meeting the ever i...

Citation Context

...) [5]–[8], [10]–[13], [16]–[18], [25], [29], [30], [33]–[35], [41], [43], [44], [97], remote storage [98], Multi-tier programming [36], [45], [46], live cloud-streaming [99], resource-aware computing =-=[100]-=-, [101], and fidelity adaptation [102] and explained as follows. Remote execution: As explained in part II-A,the resourcehungry components of mobile applications —in whole or part —are migrated to t...

Citation Context

...cture which is then mapped to the network topology. Placement algorithms follow different paradigms, e.g. iterative migration [3], limited flooding [1] and divide & conquer with iterative refinements =-=[15]-=-. We omit a discussion of these approaches due to space limitations. 4 Classification and evaluation When comparing the different approaches presented in the previous section, the most interesting asp...

Citation Context

...pensable to maximize the computational efficiency. Research in this domain, focused on mobile and embedded systems, considers limitations in battery capacities and demands for longer device lifetimes =-=[11, 28, 33, 34]-=-. In high performance and enterprise systems, power management has been driven by two significant issues. First, there are limitations in the power delivery to and cooling capabilities of large scale ...

Citation Context

...one. We want to mention that this simulation setup is very simple and does not account for communication/coordination overhead; it just considers energy savings. VI. BEYOND COOPERATIVE COMMUNICATION We have so far discussed users’ cooperation in the context of relaying/forwarding each other’s traffic; i.e., communication cooperation. Cooperation, however, can go beyond just data forwarding. Cooperation can encompass other forms of tasks, such as storage and computation. Nowadays, smartphones have powerful computational resources that can be used to cooperatively run sophisticated applications [12, 14, 15]. This can be referred to as cooperative computation. Here, each smartphone is considered as a single processing unit, and the overall environment is considered as a multi-processor environment. The computational task is then to be divided among all the nodes that are willing to cooperate. These selected nodes will exchange information among themselves using short range wireless networking. It is important, in such cases, to have a monitor component to respond to dynamic changes in application requirements and in network topologies, such as node mobility or failure. Two important components ar...

Citation Context

...) [5]–[8], [10]–[13], [16]–[18], [25], [29], [30], [33]–[35], [41], [43], [44], [97], remote storage [98], Multi-tier programming [36], [45], [46], live cloud-streaming [99], resource-aware computing =-=[100]-=-, [101], and fidelity adaptation [102] and explained as follows. • Remote execution: As explained in II-A,the resourcehungry components of mobile applications —in whole or part —are migrated to the re...

Citation Context

...come increasingly evident in both mobile and enterprise environments. Limitations in battery capacities and demands for longer device lifetimes have motivated research for mobile and embedded systems =-=[12, 34, 37, 38]-=-. For server systems, online management is important for two significant reasons. First, there are power delivery and cooling limitations in datacenter environments that result from meeting the ever i...

Citation Context

... because they are usually equipped with limited battery capacity. Some of them adopted profiling–partitioning technology to identify offloaded parts of an application for energy saving [3], [8], [9], =-=[11]-=-, [21]. They first profiled the energy consumption of each function of the application. According to the profiling result, they next generated a cost graph, in which each node represented a function t...

Citation Context

...e they usually come withsa limited battery capacity. Few of them adopted profilingspartitioning method to identify offloaded parts of ansapplication for saving energy of mobile devices [3], [4], [6],s=-=[7]-=-, [8]. Authors in [5] designed a similar method to determineswhether Java methods and byte code to native codescompilation should be executed at remote servers to savesenergy but their methods cannot ...

Citation Context

...) [5]–[8], [10]–[13], [16]–[18], [25], [29], [30], [33]–[35], [41], [43], [44], [96], remote storage [97], Multi-tier programming [36], [45], [46], live cloud-streaming [98], resource-aware computing =-=[99]-=-, [100], and fidelity adaptation [101] and explained as follows. • Remote execution: As explained in II-A,the resourcehungry components of mobile applications —in whole or part —are migrated to the re...