Recently, the release of trace data sets from several wireless network communities has stimulated much research on empirical mobility models in wireless local area networks. Many of these models have tried to extract the user mobile features from the trace data and establish the mobility model that can be used to infer or determine user mobile patterns. However, these models have ambiguously used user association patterns and user mobile patterns in realizing mobility models, where the association patterns can be perceived directly from the trace data and the user mobile patterns can be filtered from the former. The limitation of the resulting mobility model is that it depends too much on the placement of access points (APs), and thus, it is limited to represent general user mobility. In this paper, we build a dual mobility model (DMM) to describe user association patterns and mobile patterns separately, based on the observation that user association patterns do not always reflect the true user mobile patterns in a densely deployed network but they correspond to the mobile patterns in a sparsely deployed network. Specifically, we first characterized the false-positive registration patterns that result from misunderstanding multiple AP signals. We then extracted those patterns from the original trace data to construct a dense mobility model that describes the user association patterns in dense networks. With the remaining trace data (positive mobile patterns), we built a sparse mobility model that regards the association patterns as mobile patterns. In the sparse modeling, we executed user profiling to accurately characterize user mobile patterns according to user's moving coverage and speed. Finally, we validated that the DMM model with user profiling is statistically more precise than the other mobility model.

In the wireless ad hoc network, utilizing the multiple channels at Medium Access Control (MAC) layer is on of the key techniques to improve the network performance. It can be done by multi-channel MAC to utilize the channels resource as much as possible. Wireless nodes are usually powered by battery and thus are limited in power capacity. The IEEE 802.11 Power Saving Mechanism (PSM) is used to conserve energy for the ad hoc networks by allowing nodes to enter doze mode when there is no need for data exchange. In this paper, we propose a hybrid and adaptive protocol, named H-MMAC, by adopting IEEE 802.11 PSM. In H-MMAC, nodes exchange control messages on default channel to negotiate the data channel during the ATIM (Ad hoc traffic Indication Message) window. The difference of H-MMAC compared to other multi-channel MAC protocols is that the other nodes can transmit data packets on data channels based on the network traffic load. This means H-MMAC can utilize the channel resources more efficiently. The simulation results show that our proposed H-MMAC improves the network performance significantly in terms of aggregate throughput, average delay and energy efficiency.

Recently overlay/underlay framework in Cognitive Radio (CR) have been studied and demonstrated the benefits such as spectrum efficiency and channel capacity maximization. We assume Secondary Users (SUs) can choose to either acquire a dedicated spectrum with constant payment or to use a Primary User (PU) band for free. However, using PU band yields delayed transmission cost. In this paper, we apply M/M/1 queueing model with heterogeneous service rate to derive the explicit expressions for the expected delays of an arbitrary SU data packets. Based on this, the SUs need to decide whether to use the Primary User (PU) band or to acquire the dedicated band. The interaction between selfish SUs is modeled as a noncooperative game. We prove the existence and uniqueness of a symmetric Nash equilibrium, and characterize the equilibrium behavior explicitly. Then an appropriate price of the dedicated band can be defined. Numerical analysis are used to prove a high degree of accuracy for the derived expressions.

In this paper, we examine user registration patterns in empirical WLAN traces, and identify elusive associations that are abused as user movements. The examination shows that about 90% of transitions are irrelevant to actual user movements. In order to refine the elusive movements, we investigate the geographical relationships among APs and propose a filtering framework for removing them from the trace data. We then analyze the impact of the false movements on an empirical mobility model. The numerical results indicate that the proposed framework improves the fidelity of the empirical mobility model.

Code optimization of the offset assignment generated in embedded systems allows for power and space efficient systems. We propose a new heuristic that uses edge classification to commutatively transformation and optimize the assignment. We introduce concept of breakable and unbreakable edges, which assists in selecting edges for path cover and edges for commutative transformation.

Most embedded systems have limited amount of memory. In contrast, the memory requirements of code (in particular loops) running on embedded systems is significant. This paper addresses the problem of estimating the amount of memory needed for transfers of data in embedded systems. The problem of estimating the region associated with a statement or the set of elements referenced by a statement during the execution of the entire set of nested loops is analyzed. Aquantitative analysis of the number of elements referenced is presented; exact expressions for uniformly generated references and a close upper and lower bound for non-uniformly generated references are derived. In addition to presenting an algorithm that computes the total memory required, we discuss the effect of transformations on the lifetimes of array variables, i.e., the time between the first and last accesses to a given array location. A detailed analysis on the effect of unimodular transformations on data locality including the calculation of the maximum window size is discussed. The termmaximum window sizeis introduced and quantitative expressions are derived to compute the window size. The smaller the value of the maximum window size, the higher the amount of data locality in the loop.