As 100-Gb/s digital coherent systems enter commercial deployment, an effort is underway to uncover the technologies that will enable the next-generation optical fiber communication systems. We envisage that future optical transport will be software-defined, enabling flexible allocation of bandwidth resources, with dynamically adjustable per-channel data rates based on instantaneous traffic demand and quality-of-service requirements, leading to unprecedented network agility. Software-defined transponders will have the programmability to adopt various modulation formats, coding rates, and the signal bandwidth based on the transmission distance and type of fiber. Digital signal processing will become increasingly ubiquitous and sophisticated, capable of compensating all types of channel impairments, enabling advanced forward error correction coding, and performing functions previously handled poorly by optical analog hardware such as spectrum shaping and demultiplexing of optical channels.

Blind equalization typically achieves parameter optimization through cost minimization using stochastic gradient descent in both batch and adaptive algorithms. In general, stochastic descent algorithms typically require large number of iterations or long data samples to converge. The batch approach is generally based on data reuse (recycling) and re-filtering to recompute the cost gradient after each iterative parameter update, thereby causing long processing delays. In this work, we present a novel steepest descent batch algorithm that does not require data recycling. We consider the popular Constant Modulus Algorithm and the Minimum Entropy Deconvolution for normalized cumulant maximization. Both algorithms utilize 4-th order cumulants. The proposed steepest descent batch implementation of both algorithms converge rapidly in a few iterations and deliver superior performance without the delay due to data recycling and refiltering.

In this paper, we overview the fundamental principles of next-generation optical Orthogonal Frequency Division Multiple Access (OFDMA)-PON systems, with a particular focus on upstream architectures capable of achieving 10+Gb/s colorless upstream transmission. We also propose a novel OFDMA-PON architecture that satisfies these requirements and is capable of exceeding 10 Gb/s upstream transmission over a single wavelength. It is first analytically shown that optical carrier suppression at the optical network units (ONUs) combined with coherent detection at the optical line terminal (OLT) successfully eliminates both in- and cross-polarization beating noise that would otherwise be generated at the OLT and would fundamentally limit upstream transmission performance. A centralized light source OFDM-based PON architecture with source-free ONUs and OLTside coherent detection is then presented and experimentally verified to achieve 20 Gb/s/λ transmission over a class B+ optical distribution network (20 km SSMF with an additional 1:32 optical split.) By thus providing very high-speed, flexible, colorless upstream transmission, the proposed architecture is an attractive candidate for next-generation PON systems capable of cost-efficiently delivering heterogeneous services.

Wireless broadband networks, especially WiMAX networks, have emerged in the industry recently and many challenging research issues arise. In this paper, we proposed a heuristic clustering algorithm for minimum cost wireless broadband overlay network deployment Moreover, we also modified and implemented two heuristic algorithms based on classic linear programming based capacitated facility location algorithms. We analyzed the theoretical worst-case performance ratio of our algorithm and our numerical results showed that our algorithm performs much better in practical network settings

To facilitate the multi-hop communications in mobile ad hoc networks, all the nodes in the networks need to cooperate and relay packets for other nodes. Unfortunately, there exist selfish nodes that may refuse to relay packets for others in order to save their own resources such as energy. Those non-cooperative behaviors will lower down the network performance significantly. Therefore, schemes need to be developed to enforce the cooperation of selfish nodes in the multi-hop communications. Credit-based algorithms have been proposed and are considered simple and effective. However, those schemes can lead to unbalanced credit distribution. That is, some nodes in hot areas can easily gain many credits so that they stop relaying for others, while some nodes in cold areas may gain few credits so that they are excluded from the network even if they are willing to cooperate. This paper aims to provide a solution to encourage the nodes in hot areas to stay in the network and help others relay the traffic. We impose a penalty for those non-cooperative nodes in order to force them to participate the communications. To further provide incentive for cooperative nodes to remain cooperative, we propose a game played by the nodes in the same layer. The imposed penalty depends on the outcome of the game. When a node chooses cooperation, the payoff it perceives is the reward from relay minus the cost resulting from the energy loss. When a node chooses not to cooperate, the payoff is the penalty. A node will select the option that generates larger payoff. Simulation results show that our scheme can enforce the nodes in hot areas to cooperate and also encourage cooperative nodes to remain cooperative

This work presents a novel utility-based routing (UBR) algorithm for wireless ad hoc networks. In multi-hop ad hoc networks, each node may be responsible for forwarding data packets for others. Thus, some nodes may consume more energy than others, and consequently die earlier This leads to unbalanced usage of the power among the nodes, network partition, and the decrease in network lifetime. The UBR algorithm we develop takes into account not only the status of each intermediate node but also the characteristic of applications. Specifically, each application is assumed to be associated with a utility function, which maps to the quality of service (QoS). Each intermediate node sets a price to signal its willingness to participate, and then the cost for involving this intermediate node is calculated as price times the transmit power. Instead of choosing the shortest routing path, the destination node picks up the routing path that gives the maximum surplus defined as utility minus cost. Our performance study based on the PARSEC simulator shows that the proposed utility based algorithm improves the energy balance when compared with the original DSR.

As the wireless systems move to third-generation (3G) and beyond, the research on joint optimization of resource allocation for the integrated multimedia teletraffic is becoming pressing. This paper aims to set up a framework for joint resource allocation in the forward link of a multi-cell wireless multimedia system. The approach we take is based on the notion of utility. That is, each user, or each class of users, is assumed to have a utility function that describes the user's monetary satisfaction, namely utility, as a function of received quality of service (QoS). We concentrate on revenue maximization, namely, the objective is to maximize the total profit generated by selling the resource to users. In addition to their scientific merit, the results are of high value to manufacturers of cellular infrastructures and network operators alike. We characterize the optimal prices and the optimal resource partition between voice and data services for an individual cell with limited power or rate. Two-cell revenue maximization has been investigated. With limited power, not all power needs to be used when the revenue is maximized.