The Security of Wireless Sensor Networks seems to be an underdeveloped but very practical research area. With the intensive study in cryptography over the years, it is not difficult for people to think of a strong protocol to protect sensitive data. However, in Wireless Sensor Networks, many such strong protocols cannot be implemented simply because of the computational constraints on those tiny sensors. Thus, finding a less expensive yet secure protocol for sensors to communicate is somehow a challenging task. Recently we propose a method that satisfies a certain level of security concerns and is more efficient than the widely used RC5.
Type-2 computation is not new to computer scientists, which can model many contemporary computing problems such as machine learning, data mining, or problems in bioinformatics where the data size has grown to the point at which we no longer treat it as a finite object. Using the original concepts of computational complexity to analyze type-2 algorithms will overlook many properties that are difficult to model in terms of the formalism that has been very successful in type-1 computation.
Currently we focus on developing a reasonable notion of explicit type-2 complexity classes based on the computational cost of Oracle Turing Machines, a natural abstract model for type-2 computation. We motivate our type-2 complexity classes by proving type-2 analogs of the classic complexity theorems such as the Union Theorem, the Gap Theorem, and the Compression Theorem. We find that the structure of complexity classes at type-2 turns out to be quite different from its type-1 counterpart.
We also investigate the notion of optimum programs for type-2 computation. We lift the original Speed-up Theorem to type-2 and provide a new notion of optimum programs in terms of the number of oracle queries made during the courses of the computation. An easy speed-up result can be obtained if we formalized query-complexity in a naive way. However, the speed-up phenomenon is not likely to be preserved if we elaborate the notion of query-complexity carefully. What should be a proper notion of optimum programs alone this line is still not clear.