Multicasting is a remarkable technology that can effectively provide point-to-multipoint communications. The multicast communication can substantially decrease traffic in a network and thus save network resources and transmission costs. If multicasting is applied to a content delivery system, however, the transmission speed must be set to the lowest one among the available capacities of links on the multicast tree for all client terminals to receive the contents simultaneously. This type of problem is especially serious for heterogeneous networks. This paper studies effective content delivery systems for non-real-time point-to-multipoint services over heterogeneous environments and proposes an adaptive delivery system to select multicasting and store-and-forward transferring data streams. The results obtained by computer simulation show that our proposed system can reduce delivery time and that it is scalable to large networks and robust against variations in network size as well as environmental heterogeneities.

With the rapid increase in demand for mobile data, mobile network operators are trying to expand wireless network capacity by deploying wireless local area network (LAN) hotspots on which they can offload their mobile traffic. However, these network-centric methods usually do not fulfill the interests of mobile users (MUs). Taking into consideration many issues, MUs should be able to decide whether to offload their traffic to a complementary wireless LAN. Our previous work studied single-flow wireless LAN offloading from a MU's perspective by considering delay-tolerance of traffic, monetary cost and energy consumption. In this paper, we study the multi-flow mobile data offloading problem from a MU's perspective in which a MU has multiple applications to download data simultaneously from remote servers, and different applications' data have different deadlines. We formulate the wireless LAN offloading problem as a finite-horizon discrete-time Markov decision process (MDP) and establish an optimal policy by a dynamic programming based algorithm. Since the time complexity of the dynamic programming based offloading algorithm is still high, we propose a low time complexity heuristic offloading algorithm with performance sacrifice. Extensive simulations are conducted to validate our proposed offloading algorithms.

In the near future, the packet switching system will need a large switching capacity. But, large processor can not be obtained in cheap way. So, it seems better to use smaller processors in a distributed fashion to obtain a large switching capacity. If several small switching devices are interconnected, it will be possible to increase the traffic handling capacity. Depending on the topology of the network, overall processing characteristics are different. Switching networks like torus etc. have the problem, because their nodes have the transit traffic. According to the increase of the switching network size, the transit traffic becomes large and the overall traffic efficiency of the network decreases significantly. The mesh network, in which all nodes are interconnected, does not have that problem, because there is no transit traffic in any node. This paper describes the composition of mesh type distributed packet switching systems. Flexible architectures are proposed, which are based on the building block structure. A node of these architectures is composed of several switching elements, and one or two intranode switches. Using these architectures, the addition of a new node to the system can be done without any modification of the fundamental units. The experimental system has been designed and built.

Satellite-terrestrial (ST) networks, in which many nodes are interconnected by both satellite and terrestrial networks, can efficiently support multicast services. This is because satellite broadcasting is suitable for a large multicast group and a terrestrial network is suitable for a small multicast group. An ST network requires a multicast routing algorithm that can select the appropriate satellite and terrestrial routes. Conventional dynamic routing algorithms for terrestrial networks cannot construct an efficient multicast routing tree because they basically select a less-expensive route when a node is added. We have developed a dynamic routing algorithm, a virtual-cost-based algorithm, for ST networks that selects the route to use according to the multicast group size when a node is added to the group. Simulation showed that the proposed algorithm is advantageous when nodes are added to or removed from the multicast group during steady-state simulation.

In leased line services used by ISPs (Internet Service Providers) the bandwidth is fixed, but the traffic changes dynamically. Therefore, there is a necessity for ISPs to accommodate extra capacity to meet peak usage demands; many resources are not used in off-peak hours. To address this, we propose an auction method for the dynamic allocation of bandwidth to ISPs sharing backbone networks. By this method, backbone networks can be used effectively as each ISP is able to secure bandwidth according to its own policy. The Internet users can also be expected to receive good services, as it enables them to obtain information about all ISPs, such as the access fee and QoS (quality of service) provided, and to select congenial ISPs from among all ISPs according to this information. In this study, we compare a dynamic bandwidth allocation service with a leased line service (fixed allocation of bandwidth to ISPs) by using the users' utility to estimate the effectiveness of the proposed method.

Due to network users' different time-preference, network traffic load usually significantly differs at different time. In traffic peak time, network congestion may happen, which make the quality of service for network users deteriorate. There are essentially two ways to improve the quality of services in this case: (1) Network service providers (NSPs) over-provision network capacity by investment; (2) NSPs use time-dependent pricing (TDP) to reduce the traffic at traffic peak time. However, over-provisioning network capacity can be costly. Therefore, some researchers have proposed TDP to control congestion as well as improve the revenue of NSP. But to the best of our knowledge, all of the literature related time-dependent pricing scheme only consider the monopoly NSP case. In this paper, a duopoly NSP case is studied. The NSPs try to maximize their overall revenue by setting time-dependent price, while users choose NSP by considering their own preference, congestion status in the networks and the price set by the NSPs. Analytical and experimental results show that the TDP benefits the NSPs, but the revenue improvement is limited due to the competition effect.

In this paper, we point out an architecture optimization problem for networks delivering services such as Video-On-Demand or, more precisely, two intertwined problems, i.e., the storage allocation of the videos among the storage nodes of the network and the choice of the network topology. We present and investigate the properties of a genetic algorithm which can handle such problems. This algorithm, as well as a greedy heuristics and simulated annealing, are then used to derive solutions in function of link and node cost parameters in a 36-node network. The results show that genetic algorithms are an effective class of algorithms for such problems, and possibly many other topology optimization problems.

In recent years, the adoption of Software as a Service (SaaS) cloud services has surpassed that of Infrastructure as a Service (IaaS) cloud service and is now the focus of attention in cloud computing. The cloud market is becoming highly competitive owing to the increasing number of cloud service providers (CSPs), who are likely to exhibit different cloud capacities, i.e., the cloud market is heterogeneous. Moreover, as different users generally exhibit different Quality of Service (QoS) preferences, it is challenging to set prices for cloud services of good QoS. In this study, we investigate the price competition in the heterogeneous cloud market where two SaaS providers, denoted by CSP1 and CSP2, lease virtual machine (VM) instances from IaaS providers to offer cloud-based application services to users. We assume that CSP1 only has M/M/1 queue of VM instances owing to its limited cloud resources, whereas CSP2 has M/M/∞ queue of VM instances reflecting its adequate resources. We consider two price competition scenarios in which two CSPs engage in two games: one is a noncooperative strategic game (NSG) where the two CSPs set prices simultaneously and the other is a Stackelberg game (SG) where CSP2 sets the price first as the leader and is followed by CSP1, who sets the price in response to CSP2. Each user decides which cloud services to purchase (if purchases are to be made) based on the prices and QoS. The NSG scenario corresponds to the practical cloud market, where two CSPs with different cloud capacities begin to offer cloud services simultaneously; meanwhile, the SG scenario covers the instance where a more recent CSP plans to enter a cloud market whose incumbent CSP has larger cloud resources. Equilibrium is achieved in each of the scenarios. Numerical results are presented to verify our theoretical analysis.

Recently, the GMPLS controlled WSON has emerged as a promising optical transport network. In order to guarantee the optical signal transmission feature without deformation, the optoelectronic 3R regenerators still need to be sparsely placed in the network, termed as translucent networks. The growing size and complexity of the translucent network requires a transition of control plane to move from the traditional centralized model to a fully distributed architecture in the future. However, centrally designed routing, wavelength assignment, and 3R regenerator allocation approaches become unfeasible under the distributed paradigm due to the outdated and inconsistent network state information. A common solution is to accelerate the update frequency of network state, but the fundamental problem remains that the inaccurate state information is still inevitable. Furthermore, it adds a significant increase to the control traffic volume which adversely degrades the performance and scalability of the network control system. In order to mitigate the impact of having inaccurate state information on network performance in the distributed systems, a novel RWA approach is proposed in this paper, termed as routing and distributed wavelength assignment with top ranked probing wavelength set computation. In our proposal, the wavelength assignment is performed by signalling process with a set of carefully preselected probing wavelengths. This set is dynamically computed based on the resource utilization each time the network state is refreshed. The PCE module is adopted in WSON control plane to be responsible for the computation of RWA and 3R allocation. The performance of the proposed approach is studied by extensive simulations. The experiment results reveal that by employing the proposed scheme, without loss on the blocking performance the inaccuracy of the wavelength availability information can be well tolerated, and the set-up delay in lightpath provisioning can be kept at a low level.

The telecommunication network is a complex system. So, it needs an enormous computation to control a network optimally. The fuzzy control is a very useful method to deal with such a complex problem. This paper discusses an application of hierarchical fuzzy system to dynamic routing. The performance is compared with some other control methods, and it is shown that the hierarchical fuzzy system has a good performance for a wide range.

With the proliferation of IEEE 802.11 wireless local area networks, large numbers of wireless access points have been deployed, and it is often the case that a user can detect several access points simultaneously in dense metropolitan areas. Most owners, however, encrypt their networks to prevent the public from accessing them due to the increased traffic and security risk. In this work, we use pricing as an incentive mechanism to motivate the owners to share their networks with the public, while at the same time satisfying users' service demand. Specifically, we propose a “federated network” concept, in which radio resources of various wireless local area networks are managed together. Our algorithm identifies two candidate access points with the lowest price being offered (if available) to each user. We then model the price announcements of access points as a game, and characterize the Nash Equilibrium of the system. The efficiency of the Nash Equilibrium solution is evaluated via simulation studies as well.

Orthogonal frequency division multiplexing (OFDM) promises to provide the necessary boost in the core networks' capacity along with the required flexibility in order to cope with the Internet's growing heterogeneous traffic. At the same time, wavelength division multiplexing (WDM) technology remains a cost-effective and reliable solution especially for long-haul transmission. Due to the higher implementation cost of optical OFDM transmission technology, it is expected that OFDM-based bandwidth variable transponders (BVT) will co-exist with conventional WDM ones. In this paper, we provide an integer linear programming (ILP) formulation that minimizes the cost and power consumption of such hybrid architecture and then a comparison is made with a pure OFDM-based elastic optical network (EON) and a mixed line rate (MLR) WDM optical network in order to evaluate their cost and energy efficiency.

With the rapid increase in demand for mobile data, mobile network operators are trying to expand wireless network capacity by deploying wireless local area network (LAN) hotspots on to which they can offload their mobile traffic. However, these network-centric methods usually do not fulfill the interests of mobile users (MUs). Taking into consideration many issues such as different applications' deadlines, monetary cost and energy consumption, how the MU decides whether to offload their traffic to a complementary wireless LAN is an important issue. Previous studies assume the MU's mobility pattern is known in advance, which is not always true. In this paper, we study the MU's policy to minimize his monetary cost and energy consumption without known MU mobility pattern. We propose to use a kind of reinforcement learning technique called deep Q-network (DQN) for MU to learn the optimal offloading policy from past experiences. In the proposed DQN based offloading algorithm, MU's mobility pattern is no longer needed. Furthermore, MU's state of remaining data is directly fed into the convolution neural network in DQN without discretization. Therefore, not only does the discretization error present in previous work disappear, but also it makes the proposed algorithm has the ability to generalize the past experiences, which is especially effective when the number of states is large. Extensive simulations are conducted to validate our proposed offloading algorithms.

We propose a new multicast ATM switching network, called the Cascade Clos Broadcast Switching Network (CCB-SN). It has the following features. Firstly, the CCB-SN is robust not only for unicast calls but also for multicast calls. Therefore, deterioration of traffic handling performance due to a non-uniform traffic pattern does not occur. Secondly, in the CCB-SN, the cells are replicated after they are partly routed. And so, the delay time caused by the increment of replicas of cells is kept to a minimum.

A store and fordward speech interpolation telephonic concentrator system is introduced. This paper presents first the system, describes its main parts and components and using typical parameters the probable system operation is outlined. Following, a study on capacity and delay of transmission of each voice packet is done by supposing that a M/D/1 model can be applied to the system. As voice has special characteristics and packets don't arrive randomly as supposed in the M/D/1 model, an analysis using a computer simulation is done. The source generating speech is closely matched with human voice by using a model with good resolution specially for small pauses of talkers during the active generation of speech. Using this source, simulations representing 120 or 180 sec. of actual time are done. As its results are different of those predicted by the M/D/1 model, the capacity of the system is forecasted again through examples. To increase even more the capacity without impairing the frozen out fraction of the speech or increasing the delay of each packed, a different method to decrease congestion in those short moments of high arrival rate of packets is intended. This method consist in to transmit from each voice sample only 7 bits during the high congestion moments, creating new capacity by shortening packets. Results of parameters obtained by simulation and probable capacity of the system are again shown.

Multicast ATM switch is in great demand for the future communication network. We have proposed the Batcher banyan network with cell copy preparation stages as a transit switch. It performs cell replication with small hardware increase. On the trunkline, multicast traffic is quite little, thus hardware for cell copy can be sustained small. In those previous works, the effect of the external blocking was omitted. In this paper, we propose a multicast switching network which adopts the incomplete copy network that we have proposed, and examine several strategies to prevent the external blocking for this switching network. Namely, the input buffer method with an arbitration network is applied. For multicast usage, we propose two modifications. One is to arbitrate after cell replication for the sake of simple control and small hardware. The other is to annex a cell distribution network for smoothing biased cell arrival. Biased cell arrival occurs because the output of the incomplete copy network is not uniform. Simulation results show the effectiveness of the proposed method.

In wireless multi-hop networks such as ad hoc networks and sensor networks, backoff-based opportunistic routing protocols, which make a forwarding decision based on backoff time, have been proposed. In the protocols, each potential forwarder calculates the backoff time based on the product of a weight and global scaling factor. The weight prioritizes potential forwarders and is calculated based on hop counts to the destination of a sender and receiver. The global scaling factor is a predetermined value to map the weight to the actual backoff time. However, there are three common issues derived from the global scaling factor. First, it is necessary to share the predetermined global scaling factor with a centralized manner among all terminals properly for the backoff time calculation. Second, it is almost impossible to change the global scaling factor during the networks are being used. Third, it is difficult to set the global scaling factor to an appropriate value since the value differs among each local surrounding of forwarders. To address the aforementioned issues, this paper proposes a novel decentralized local scaling factor control without relying on a predetermined global scaling factor. The proposed method consists of the following three mechanisms: (1) sender-centric local scaling factor setting mechanism in a decentralized manner instead of the global scaling factor, (2) adaptive scaling factor control mechanism which adapts the local scaling factor to each local surrounding of forwarders, and (3) mitigation mechanism for excessive local scaling factor increases for the local scaling factor convergence. Finally, this paper evaluates the backoff-based opportunistic routing protocol with and without the proposed method using computer simulations.

Many new multimedia applications involve multiple dynamically changing participants, have stringent source-to-end delay requirements, and consume large amounts of network resources. A conventional algorithm that allows "two coming paths," where nodes in a multicast tree transmit several identical data flows, is therefore not practical. We have developed an algorithm for delay-constrained dynamic routing. This algorithm uses a QoS label to prevent the occurrence of "two coming paths," and can construct an efficient multicast tree for any traffic volume. The proposed algorithm was superior to conventional routing algorithms in terms of cost when nodes were added to or removed from the multicast group during a steady-state simulation.