Virtual machine(VM)-based computation offloading suffers from long delay for VM provisioning. Existing researches are not suitable for edge computing since they are designed for resource rich cloud servers. In this paper, we reduce the delay by pre-provisioning and cloning them later. We collect a large-scale measurement trace of global users and combine them with the existing Google cloud trace. Extensive evaluation indicates that the proposed scheme outperforms the existing schemes in edge nodes.

Even if a collision occurs in IEEE 802.11 network, a transmission may be successfully decoded at the receiver if the signal strength of one transmission is sufficiently stronger than the other transmission. This phenomenon is called “Physical Layer Capture” (PLC). While existing works have considered PLC between data frames, in this paper we investigate the case that an ACK frame collides with the unfinished transmission of other data frames after the occurrence of PLC between data frames. As a result of this collision, the ACK frame may be corrupted and the corresponding data frame needs to be retransmitted. We call this phenomenon “ACK Corruption” (AC). We identify the characteristic of AC via extensive experiments and simulations. Our study reveals that AC can occur in all IEEE 802.11 variants and its chance is dependent upon the relative signal strength between the stations and the MCS setting used. Further, we devise a way to avoid AC occurrence and evaluate its effectiveness.

While wearable devices are equipped with small battery, the amount of data that they send and receive has already exceeded the capacity of the energy efficient link technologies (e.g., Bluetooth Low Energy). Existing higher capacity link (e.g., Bluetooth Classic, WiFi) however requires high connection setup cost. To deal with this dilemma, we propose to bundle multiple data transmission requests and send the aggregated data at once to reduce energy consumption for link setup. Where the concept of bundling has been widely used for cloud storage services, existing schemes do not consider energy efficiency. In this paper, we present three bundling algorithms that try to balance the reduction of energy consumption and the extension of delivery delay by adapting to the current data request pattern. We perform extensive evaluation by using the realistic parameters that are obtained via experimental measurements by using a latest commercial smart watch.

Physical layer capture is one of the main causes of unfairness in IEEE 802.11 Wireless LANs. While existing work have analyzed the impact of physical layer capture on the total system throughput of a cell, we analyze the unfairness among users as result of capture. Since this unfairness is related to the relative location of users, we call it as spatial unfairness. Our analysis characterizes the relative throughput share of a user as a function of the users' distance from the Access Point (i.e., as a function of received signal strength). We validate the accuracy of our analysis via extensive simulations and experimental measurements.

Energy efficiency is one of the most important issues of Wireless sensor networks (WSN) due to limited battery capacity. In sink-centric WSN (i.e., sensing data is collected by the sink node), this issue is particularly challenging due to the `funnel effect' which means high energy consumption of the nodes near the sink. While clustering can be used to achieve scalability of WSN, previous works on energy optimal design of the cluster structure have limitations in the scalability. In this paper, we propose a scheme to build an energy optimal cluster structure for a sink-centric WSN. The most important finding of this paper is that energy optimality depends only the size of first three clusters from the sink node. Based on this finding, we design a very scalable clustering algorithm for optimal network lifetime. The algorithm is highly scalable since its time complexity is independent of the network size. We verify the correctness of the algorithm by comparing with the optimal results that are found by exhaustive search. Numerical evaluations are performed by modeling the off-the-shelf radio links such as IEEE 802.15.4 and IEEE 802.11.

Seamless is an essential feature of cellular networks. For small-cell networks, effective becomes particularly challenging. If some cells may be activated and deactivated dynamically, effective handling will become even more difficult. A key factor of good performance is handover timing, that is, making decision at a right time. If is executed too early or too late, users will experience temporary link disconnection, called radio link failure RLF. Handover timing is controlled by parameters, which are set by the network. RLF prevention is directly related to the proper configuration of these parameters. In this paper, we propose a self-optimization scheme that adjusts the parameters to minimize RLFs for dynamic small-cell networks. Our scheme first detects the types of RLF and then adjusts the parameters according to the types of RLF. Unlike most existing schemes, our scheme adjusts both system common parameters and cell-specific parameters together. In certain situations, such as when two adjacent cells do not have sufficient coverage overlap, RLFs may not be reduced to a satisfactory level by the parameter adjustment alone. To deal with such a case, our scheme adjusts the cell coverage in conjunction with the parameter optimization. The simulation results show that the proposed scheme can virtually eliminate RLFs. Copyright © 2013 John Wiley & Sons, Ltd.

Knowing the available throughput of the IEEE 802.11 link is highly beneficial. One use case of such information is the selection of an Access Point (AP) that will provide highest throughput. In this paper, we present a scheme for non-intrusive estimation of expected throughput of an 802.11 device. The proposed scheme does not require any aid from the infrastructure (e.g., AP) and does not create any additional traffic overhead (e.g., probe traffic). Our scheme relies on passive sniffing of the ongoing traffic for short duration (typically less than a second). Via sniffing, we obtain necessary parameters for throughput estimation, such as the collision probability. We have implemented our scheme in Linux system and Android smartphone. Experimental evaluation shows that our scheme accurately estimates expected throughput in various experiment settings.

Physical layer capture is one of the main causes of throughput unfairness in IEEE 802.11 Wireless LANs. While existing work have analyzed the impact of physical layer capture on the total system throughput of a cell, we study the relative unfairness among users as result of capture. Since this unfairness is related to the relative location of users, we call it as spatial unfairness. Our evaluation characterizes the relative throughput of users as a function of the users' distance from the Access Point.

A number of routing schemes have been proposed for intermittently connected mobile networks. In these schemes, a node carries a message and forwards the message to another node that it meets. To increase the delivery success rate, typically, multiple copies of the message are created and forwarded to multiple neighbor nodes. The neighbor nodes are selected based on such criteria as the probability that the neighbor node is likely to meet the destination. However, the risk of forwarding failure due to link disconnection has not been considered. In this paper we consider forwarding failure in the forwarding decision. We first develop a mathematical model to catch an optimal fragmentation size of message that minimizes the message delivery delay while achieving a certain level of message delivery ratio. Then we propose an real-time decision algorithm which decides during routing whether a certain neighbor node is appropriate for message forwarding.

Network lifetime maximization is an important problem particularly for large-scale wireless sensor networks. To this end, we take an approach that exploits the `sink mobility'. The sensor nodes near the sink node tend to suffer high energy consumption due to heavy traffic relay operations. As the sink node moves around, such energy bottleneck can be alleviated. We analyze the optimal sink mobility pattern obtained by solving a linear programming model and use the insights from the analysis to design a simple practical heuristic. Our heuristic utilizes only local information, e.g., the standard deviation of residual energy of neighboring sensor nodes. The simulation results show that our scheme achieves near-optimal performance even with the slowly moving mobile sink.

For large-scale wireless sensor networks (WSNs), cluster-based architecture is desirable. Conventional short range sensor radio, however, is often inefficient for the inter-cluster communication due to the large distance between cluster heads. One solution is to use a longer range radio for inter-cluster head communication. In this paper, we consider a dual-radio cluster head with IEEE 802.15.4 and IEEE 802.11b. A key issue is that 802.15.4 and 802.11b use the same frequency band at 2.4 GHz, so that they can potentially interfere with each other. We analyze this interference behavior via testbed experiments. Our results indicate that with a simple relaying mechanism, interference was quite high even when orthogonal channels are used for two radios. This strongly suggests the need of smarter relaying mechanisms.

In wireless sensor networks (WSN), energy efficiency is critical, particularly for large-scale deployments. In this paper, we analyze the benefits of using dual-radio, from the perspective of network lifetime, in conjunction with clustering architecture, which is widely adopted for scalability. Two types of radios, which have different property of energy consumption, are used for different purposes. We formulate a condition that warrants the benefits of dual-radio, and evaluate the effectiveness of dual-radio via simulations.

Wireless networks have been evolved to heterogeneous overlay structure, in which various wireless technologies provide partial or full coverage of the service area. Another evolution direction is multi-service wireless networks. In multi-service networks, traffic of various QoS classes is serviced concurrently. In this paper, we propose a scalable and flexible network selection scheme for multi-service heterogeneous wireless networks. In the proposed scheme, the profiles are generated for users or cells to represent their characteristics in three categories, which are traffic status, mobility and QoS. The best cell is determined for each user via profile matching. The proposed scheme can be tailored to utilize the information available from the underlying wireless technologies. Via extensive simulations, it is shown that the proposed scheme handles the interplay between multi-service traffic classes more effectively than existing schemes. We discuss how the proposed scheme can be applied to the 3GPP LTE networks.

Seamless mobility support and efficient network resource utilization is the key to the performance of mobile multimedia services. Currently, the PMIP (Proxy Mobile-IP) and the IMS (IP Multimedia Subsystem) are considered as the primary means to this end, mostly focusing on unicast applications (e.g., voice over IP). In this paper, we extend the application scope to multicast services and propose a solution framework combining the PMIP with IP multicast. Specifically, we propose an extension of the PMIP (called PMIP-M) for a seamless handover of multicast services that use existing IP multicast protocols, where gradual IP multicast deployment is taken into account for practicality. That is, a user can start or continue to receive multicast service even when he migrates to a network without IP multicast capability. The performance of the proposed scheme is analyzed by extensive simulations.

Physical layer capture is one of the basic causes of throughput unfairness in IEEE 802.11 Wireless LANs. While papers have analyzed the impact of capture on the overall throughput of a single 802.11 cell, we are unaware of any analysis of the relative unfairness among users as result of capture. Since this unfairness is related to the relative location of users, we call this as spatial unfairness. We provide, to the best of our knowledge, the first such analysis that characterizes the relative throughput of users as a function of the users' distance from the access point (AP). We experimentally validate our analysis using a single 802.11b cell testbed.

Maximizing network throughput while providing fairness is one of the key challenges in wireless LANs (WLANs). This goal is typically achieved when the load of access points (APs) is balanced. Recent studies on operational WLANs, however, have shown that AP load is often substantially uneven. To alleviate such imbalance of load, several load balancing schemes have been proposed. These schemes commonly require proprietary software or hardware at the user side for controlling the user-AP association. In this paper we present a new load balancing technique by controlling the size of WLAN cells (i.e., AP's coverage range), which is conceptually similar to cell breathing in cellular networks. The proposed scheme does not require any modification to the users neither the IEEE 802.11 standard. It only requires the ability of dynamically changing the transmission power of the AP beacon messages. We develop a set of polynomial time algorithms that find the optimal beacon power settings which minimize the load of the most congested AP. We also consider the problem of network-wide min-max load balancing. Simulation results show that the performance of the proposed method is comparable with or superior to the best existing association-based methods.

In this paper, we propose a novel load balancing routing algorithm for wireless mesh networks (WMNs) with multiple gateways, i.e., sink routers. As most traffic occurring in WMNs is relayed to these gateways through wireless links in a multi-hop fashion, the sink routers are easily overloaded. In such a situation, when a user initiates a communication session and the corresponding route toward the sink router cannot provide the required level of services, the session is rejected. Our objective is to achieve load balancing routing that minimizes such rejections. To this end, we take an adaptive partitioning approach, which heuristically divides the entire network into partitions and applies load balancing to between partitions and within partitions. To cope with user mobility, partitions and routes are adaptively managed. Evaluation via simulations indicates up to 25% performance gains over existing schemes in rejected requests, while the overhead in route computation costs drops by up to 27%.

Energy-aware routing is important in multi-hop wireless networks that are operated by battery power, e.g., wireless sensor networks. To maximize the network survivability, the energy efficiency of paths must be taken into account for route selection. Simple heuristics such as choosing paths with minimal energy consumption do not perform well, because the energy of the nodes on such paths may deplete quickly. The issue is particularly serious for the networks with regular traffic pattern as in monitoring sensor applications. Existing solutions to this issue typically adopt the multi-path routing approach, in which multiple paths are set up between source and destination and one (or all) of the paths is (are) used at a certain moment. However, this approach involves high overhead for establishment and management of multiple paths. In this paper, we present a static single-path routing algorithm which uses one energy-efficient path for each communicating peer throughout the network lifetime, eliminating the overhead of multi-path routing. It is proved that our algorithm achieves a constant factor approximate of the optimal solution. We compare the performance of the proposed scheme with that of multi-path routing via simulations. Despite the use of single static path, the proposed scheme outperforms the dynamic multi-path approach, particularly under heavy network loads.

Maximizing network throughput while providing fairness is one of the key challenges in wireless LANs (WLANs). This goal is typically achieved when the load of access points (APs) is balanced. Recent studies on operational WLANs, however, have shown that AP load is often substantially uneven. To alleviate such imbalance of load, several load balancing schemes have been proposed. These schemes commonly require proprietary software or hardware at the user side for controlling the user-AP association. In this paper we present a new load balancing technique by controlling the size of WLAN cells (i.e., AP’s coverage range), which is conceptually similar to cell breathing in cellular networks. The proposed scheme does not require any modification to the users neither the IEEE 802.11 standard. It only requires the ability of dynamically changing the transmission power of the AP beacon messages. We develop a set of polynomial time algorithms that find the optimal beacon power settings which minimize the load of the most congested AP. We also consider the problem of network-wide min-max load balancing. Simulation results show that the performance of the proposed method is comparable with or superior to the best existing association-based methods.

Recent studies on operational wireless LANs (WLANs) have shown that user load is often unevenly distributed among wireless access points (APs). This unbalanced load results in unfair bandwidth allocation among users. We observe that the unbalanced load and unfair bandwidth allocation can be greatly alleviated by intelligently associating users to APs, termed association control, rather than having users greedily associate APs of best received signal strength.In this study, we present an efficient algorithmic solution to determine the user-AP associations that ensure max-min fair bandwidth allocation. We provide a rigorous formulation of the association control problem that considers bandwidth constraints of both the wireless and backhaul links. Our formulation indicates the strong correlation between fairness and load balancing, which enables us to use load balancing techniques for obtaining near optimal max-min fair bandwidth allocation. Since this problem is NP-hard, we present algorithms that achieve a constant-factor approximate max-min fair bandwidth allocation. First, we calculate a fractional load balancing solution, where users can be associated with multiple APs simultaneously. This solution guarantees the fairest bandwidth allocation in terms of max-min fairness. Then, by utilizing a rounding method we obtain an efficient integral association. In particular, we provide a 2-approximation algorithm for unweighted greedy users and a 3-approximation algorithm for weighted and bounded-demand users. In addition to bandwidth fairness, we also consider time fairness and we show it can be solved optimally. We further extend our schemes for the on-line case where users may join and leave. Our simulations demonstrate that the proposed algorithms achieve close to optimal load balancing and max-min fairness and they outperform commonly used heuristic approaches.

Hierarchical wireless overlay networks have been proposed as an attractive alternative and extension of cellular network architectures to provide the necessary cell capacities to effectively support next-generation wireless data applications. In addition, they allow for flexible mobility management strategies and quality-of-service differentiation. One of the crucial problems in hierarchical overlay networks is the assignment of wireless data users to the different layers of the overlay architecture. In this paper, we present a framework and several analytical results pertaining to the performance of two assignment strategies based on the user's velocity and the amount of data to be transmitted. The main contribution is to prove that the minimum average number of users in the system, as well as the minimum expected system load for an incoming user, are the same under both assignment strategies. We provide explicit analytical expressions as well as unique characterizations of the optimal thresholds on the velocity and amount of data to be transmitted. These results are very general and hold for any distribution of user profiles and any call arrival rates. We also show that intelligent assignment strategies yield significant gains over strategies that are oblivious to the user profiles. Adaptive and on-line strategies are derived that do not require any a priori knowledge of the user population and the network parameters. Extensive simulations are conducted to support the theoretical results presented and conclude that the on-line strategies achieve near-optimal performance when compared with off-line strategies.

One of the crucial problems in hierarchical overlay networks is the assignment of users to the different layers of the overlay architecture. In an earlier paper [T.E. Klien and S.J. Han, March 12-14, 2003], we have investigated the system-wide performance of assignment strategies based on the users' profile information of velocity and amount of data to be transmitted. The optimal assignment strategies that minimize the average number of users in the system rely on knowledge of the cell capacities, the call arrival rates and the statistical description of the users' profiles. In realistic operating conditions, this knowledge may not be available and would typically be time-varying. In this paper, we propose adaptive and online assignment strategies that do not require any a priori knowledge, but still achieve the same performance as if such knowledge were available.

The third-generation (3G) wide area wireless networks and 802.11 local area wireless networks possess complementary characteristics. 3G networks promise to offer always-on, ubiquitous connectivity with relatively low data rates. 802.11 offers much higher data rates, comparable to wired networks, but can cover only smaller areas, suitable for hot-spot applications in hotels and airports. The performance and flexibility of wireless data services would be dramatically improved if users could seamlessly roam across the two networks. In this paper, we address the problem of integration of these two classes of networks to offer such seamless connectivity. Specifically, we describe two possible integration approaches - namely tight integration and loose integration and advocate the latter as the preferred approach. Our realization of the loose integration approach consists of two components: a new network element called IOTA gateway deployed in 802.11 networks, and a new client software. The IOTA gateway is composed of several software modules, and with cooperation from the client software offers integrated 802.11/3G wireless data services that support seamless intertechnology mobility, Quality of Service (QoS) guarantees and multiprovider roaming agreements. We describe the design and implementation of the IOTA gateway and the client software in detail and present experimental performance results that validate our architectural approach.