Recently, the topic for the development of efficient handover algorithms to select the target (serving) base station (BS) such that the average throughput of user equipments (UEs) is maximized in heterogeneous cellular networks has received much attention. In the literature, the latest handover algorithms employ the upper confidence bound (UCB) policy, one efficient method for the multi-armed bandit (MAB) problem, since the UCB policy can identify the critical actions quickly by taking the uncertainty of the action reward into account. However, the UCB policy does not take advantage of the observed features, which could lead to poor performance in the UE throughput since the throughput of a UE served by a BS is highly correlated to the observed features including the bandwidth of the BS, the number of UEs served by the BS, and the spectral efficiency of the UE achieved by the BS. By contrast, besides the consideration of the uncertainty of the action reward, the linear upper confidence bound (LinUCB) policy also makes use of the observed features to estimate the action reward. In this paper, we make the first attempt to propose a LinUCB-based handover algorithm. The challenge is to define the reward and observed features highly correlated to the UE throughput such that the reward is a linear combination of the observed features as demanded by the LinUCB policy. We conduct simulations on the network simulator ns-3 to show that the proposed algorithm outperforms the state-of-the-art handover algorithms in terms of the average UE throughput.

In the Third Generation Partnership Project (3GPP) specification [1], the conditional hand over procedure is defined to reduce the hand over failures in 5G networks by early preparation of the radio resource for user equipment (UE). In this paper, in order to minimize the resource reservation time while maintaining the radio link failures (RLF) rate of a UE, we make the first attempt to trigger the conditional handover of the UE only when the UE would experience an RLF in the near future. Our idea is to adjust the handover margin of triggering the conditional hand over of a UE based on the prediction on whether the UE would experience an RLF in the near future. To achieve an accurate prediction in the urban scenario with blockages, we reduce the conditional handover problem to a classification problem and solve the classification problem using a recurrent neural network (RNN). Simulations show that the proposed algorithm has extraordinary performance in terms of the RLF rate and resource reservation time, as compared with the state-of-the-art methods.

A traditional 4G handover algorithm that performs well in a macro-cell-only network could not be employed in the 5G network with the random distribution of small base stations due to the irregular change of the signal-to-interference-plus-noise ratio (SINR) of a moving user equipment (UE). Besides, the fuzzy logic is a well-known method to translate the domain knowledge of a human expert into a set of basic rules and to formalize the uncertainty judged by the human expert. In this paper, based on the fuzzy logic, we make the first attempt to propose a handover algorithm for a UE in 5G networks. Simulations show that our algorithm has a good performance in terms of the radio link failure (RLF) rate and the ping-pong rate in a 5G network, as compared with the state-of-the-art methods.

Multi-protocol label switching (MPLS) is supported by OpenFlow version 1.2 or higher and widely used in software defined networking (SDN) to achieve higher performance and flexibility. Due to the shortage of the ternary content addressable memory (TCAM), the number of forwarding entries installed in a switch needs to be bounded (node path-degree constraints). Besides, the maximum concurrent flow problem, which asks to maximize the minimum fraction of the flow of a commodity to its demand, is widely studied because of a wide range of applications. In this paper, we address the maximum concurrent flow problem while ensuring the flow routed for a commodity does not exceed its demand (demand constraints) and the node path-degree constraints are imposed, termed the bounded path- degree maximum concurrent flow (BPMCF) problem. We first show the BPMCF problem is NP-hard and intractable to devise any approximation algorithm. Then, we propose an algorithm for the BPMCF problem. Finally, we evaluate the performance of the proposed algorithm through computer simulations and experiments on Global Environment for Network Innovations (GENI) testbed using the real-life traces collected from SNDlib.

In this paper, we study a multiple sink placement problem with latency and reliability guarantee in a lossy wireless sensor network and its applications in an advanced metering infrastructure (AMI) network. We show the problem is NP-hard and propose an algorithm for the problem. The proposed algorithm jointly considers the routing protocol RPL [1], the MAC protocol TSCH [2], and the scheduling approach DeTAS [3] to meet the latency and reliability requirements in a real AMI network. We conduct simulations on the real data for the establishment of an AMI network. Simulation results show the proposed algorithm has a good performance in the number of selected concentrators and the balanced number of smart meters connected to the selected concentrators.

The notion of diagnosability has long played an important role in measuring the reliability of multiprocessor systems. Such a system is t-diagnosable if all faulty nodes can be identified without replacement when the number of faults does not exceed t, where t is some positive integer. Furthermore, a system is strongly i-diagnosable if it can achieve (t + 1)-diagnosability, except for the case where a node's neighbors are all faulty. In this paper, we investigate the strong diagnosability of a class of product networks, under the comparison diagnosis model. Based on our results, we can determine the strong diagnosability of several widely used multiprocessor systems, such as hypercubes, mesh-connected k-ary n-cubes, torus-connected k-ary n-cubes, and hyper-Petersen networks.

A system is t-diagnosable if all faulty nodes can be identified without replacement when the number of faults does not exceed t, where t is some positive integer. Furthermore, a system is strongly t-diagnosable if it is t-diagnosable and can achieve (t+1)-diagnosable except for the case where a node's neighbors are all faulty. In this paper, we propose some conditions for verifying whether a class of interconnection networks, called matching composition networks (MCNs), are strongly diagnosable under the comparison diagnosis model.