Abstract: A communication path is established that allows for communication between a filter graph space and a network driver space in a system. The communication path allows bi-directional communication between a filter graph driver in the filter graph space and a network driver in the network driver space. Addresses of data can be passed from the network driver to the filter graph driver, which allows the filter graph driver to communicate with a filter graph manager as necessary to configure the filter graph space to obtain data from the requested address. The data obtained from the address is then passed from the filter graph driver to the network driver.

Abstract: A method for generating random bits which is as fast as popular generators and which has provable strong properties. The method makes use of the unpredictability properties of a relatively slow cryptographically strong generator and the rapid mixing properties of random walks on expander graphs. Use of the cryptographically strong generator is typically restricted to a small off-line or a preprocessing step. Subsequent bits are produced with a few register operations per bit. The method and concomitant generator are useful in at least two different types of applications: for faster reliable simulations (or randomizing algorithms), and cryptographic/encryption schemes.

Abstract: Methodology and concomitant circuitry for compacting an incoming data stream into an outgoing compacted data stream utilize a plurality of memories or lists. The incoming data stream is partitioned into a sequence of tokens. A primary memory stores each token, with the most recently appearing token occupying the top rank in the list. A secondary memory stores the location in the primary memory of each of a subset of tokens. The compacted data stream is generated as a coded representation of the token itself, the position of the token in the primary memory, or the position in secondary memory of the location of the token in primary memory. A tertiary list may also be employed to generate a coded representation of the position in the tertiary list of the secondary list. Searching of the lists is effected with a hashing function. Updating of the lists utilizes tree splaying.