Abstract: Various embodiments are directed toward compressing and/or decompressing data communicated between one or more network devices (e.g., codec operations). In particular, embodiments are directed towards improving codec performance by seeding the computation workspace that may be used by various codec processors. The seeding data may be determined based on at least one characteristic of a particular codec and the characteristics of data that may be processed by the codec processor. Also, the codec processor may be employed to generate data for the codec workspace based on the determined seeding data. Workspace data may be generated by processing the seeding data with the same codec processor that is used for normal codec operations. The workspace generated from the seeding data may be stored for future use, such as, when a matched data stream arrives.

Abstract: Embodiments compress and encrypt data in a single pass to reduce inefficiencies that occur from compression and encrypting data separately. Typically, compression and encryption are implemented in separate functional units. This has a few disadvantages: 1) encryption cannot make use of compression state to further secure the message, 2) processed data is read and written twice, 3) additional space, time, and resources are consumed, and 4) it is more prone to potential cipher-attacks since the encryption stage is independent from compression. Embodiments overcome these disadvantages by structuring these operations so that both compression and encryption is executed within the same processing loop. Thus: 1) encryption is stronger due to the dependence on the compression state, 2) I/O buffers are accessed only once reducing overhead, 3) system footprint is reduced, and 4) cipher analysis is more complex since the decryption process cannot be separated from the decompression process.