It is cost-effective to install multiple fibers in each link of an all-optical network, because the cost of fibers is relatively low compared with the installation cost. The resulting network can provide a large capacity for good quality of service, future growth, and fault tolerance. If a node has more incoming/outgoing fibers, it requires larger optical switches. Using the current photonic technology, it is difficult to realize large optical switches. Even if they can be realized, they are expensive. To overcome this problem, we design a node configuration for all-optical networks. We exploit the flexibility that, to establish a lightpath across a node, we can select any one of the available channels in the incoming link and any one of the available channels in the outgoing link. As a result, the proposed node configuration requires significantly smaller optical switches while it can result in nearly the same blocking probability as the existing one. We demonstrate that a good network design is to adopt the proposed node configuration and slightly more fibers in each link, so that the network requires small optical switches while it has a small blocking probability.

We design and optimize the physical topology of all-optical networks. This problem is more challenging than the traditional one for electronic communication networks, because of the wavelength-continuous constraint and it involves routing and wavelength assignment. In this problem, we are given the number of lightpaths required by every node pair and a cost specification, and our objective is to determine a physical topology of minimal cost. We formulate the problem, prove that it is NP-hard, and design an efficient algorithm called two-stage cut saturation algorithm for it. In the first stage, we relax the wavelength-continuous constraint and apply the main idea of the cut saturation method to determine a good initial network. In the second stage, we impose the wavelength-continuous constraint and perform routing and wavelength assignment to establish the specified lightpaths on the initial network. When some lightpaths cannot be established, we apply the main idea of the cut saturation method to optimize the insertion of additional links into the network. Simulation results show the following: (1) the proposed algorithm can efficiently design networks with low costs and high utilization and (2) if wavelength converters are available to support full wavelength conversion, the total cost of the links can be significantly reduced.

The paper investigates image retrieval methods based on color, shape and spatial analysis. We have designed and implemented a prototype to retrieve a particular image from an image database. Indexing methods based on criteria are adopted in the prototype. An integrated method for calculating the similarity value between two images is introduced. We then evaluate the performance and compare the characteristic of each image retrieval approach in turn. Experimental results show that using color-only analysis can achieve a good image recall rate. Finally, integrating the color, shape and spatial analysis in image retrieval, the accuracy is much higher than using the techniques separately.

Many Chinese input methods rely heavily on the dictionary stroke orders of Chinese characters. Any deviation from the standard stroke orders in these methods would result in incorrect inputs. We propose a new coding scheme for handprinted Chinese characters, which allows stroke-order free input methods to be derived. In this scheme, each handprinted Chinese character is assigned an n-or-less-digit boxing code, which is derived from the box-like structure of Chinese characters. Each digit of the boxing code represents a unique stroke type of a given character stroke. Subsequent character recognition is then based on the matching of these boxing codes. Experiments were conducted based on 6 basic stroke types (t=6), and a boxing code length of 10 (n=10). Of the 300 characters we have tested, the boxing codes generated from both the dictionary and random stroke orders were the same for each character. The duplication rate was 5.6%. The results of the test indicate that boxing code is a feasible representation of Chinese characters. Although duplications exist, these could be reduced, if not totally removed, by taking in more stroke types. Further investigation is worthwhile in the fine-tuning of the two parameters t and n so as to obtain a more accurate and wider representation, and in the application of the new coding scheme.

Multihop CDMA wireless networks with a large number of stations would require a large number of spreading codes, even with code reuse beyond the interference range. In this paper, we propose the least conflict sharable code assignment algorithm to reduce the number of codes needed and at the same time avoid code collisions as far as possible. We also introduce the mean conflictness of a network as a performance index of code assignment algorithms. Numerical results show that this new algorithm can reduce nearly 20 percent of the mean conflictness when compared to randomly sharing the codes.

A protocol called the simple tone sense (STS) protocol is designed for multihop packet radio networks (PRNs) with multiple directional antennas stations. The protocol can minimize transmission interference by using a group of tones to identify the active neighbors. A variation of the STS protocol called the variable power tone sense (VPTS) protocol is also designed to further reduce interference. Algorithms for assigning tones and for determining the orientation and broadcasting angles of the directional antennas are designed. Design examples are given. Simulation result shows that the STS protocol gives better throughput-delay performance than the busy-tone multiple access protocol, especially when the traffic is heavy. The VPTS protocol gives still better throughput-delay performance than the STS protocol.>