公开(公告)号：US20150281311A1

公开(公告)日：2015-10-01

申请号：US14673420

申请日：2015-03-30

Applicant: Comtech EF Data Corp.

Inventor： Seth Sjoholm ,  Edward Ray Coulter

IPC: H04L29/06 ,  H04L29/08

CPC classification number: H04L69/04 ,  H03M7/30 ,  H03M7/40 ,  H04L29/08072

Abstract: A method of decoding a dynamic Huffman encoded data stream comprising receiving, by a processor, one or more input bits of encoded data from among an encoded data stream, searching, by the processor, a ternary content addressable memory (TCAM) for one or more codewords, accessing, by the processor, a plurality of random access memories (RAMs) that comprise the TCAM and have a maximum number of entries that is less than or equal to one half of two to a power of a maximum number of bits to be searched, reading a value from a first level RAM and using the value to index to a second level RAM, and outputting, by the processor, a codeword length and a decoded codeword value from either the first or second level RAM.

Abstract translation: 一种解码动态霍夫曼编码数据流的方法，包括由处理器从编码数据流中接收编码数据的一个或多个输入位，由处理器检索一个或多个三进制内容可寻址存储器（TCAM） 代码字，由处理器访问包括TCAM并且具有小于或等于二分之一的最大数量的条目的多个随机存取存储器（RAM），以达到最大位数的功率 搜索，从第一级RAM读取值并使用该值索引到第二级RAM，并且由处理器从第一级或第二级RAM输出码字长度和解码码字值。

公开(公告)号：US09832287B2

公开(公告)日：2017-11-28

申请号：US14673420

申请日：2015-03-30

Applicant: Comtech EF Data Corp.

Inventor： Seth Sjoholm ,  Edward Ray Coulter

IPC: H03M7/40 ,  H04L29/06 ,  H03M7/30 ,  H04L29/08

CPC classification number: H04L69/04 ,  H03M7/30 ,  H03M7/40 ,  H04L29/08072

Abstract: A method of decoding a dynamic Huffman encoded data stream comprising receiving, by a processor, one or more input bits of encoded data from among an encoded data stream, searching, by the processor, a ternary content addressable memory (TCAM) for one or more codewords, accessing, by the processor, a plurality of random access memories (RAMs) that comprise the TCAM and have a maximum number of entries that is less than or equal to one half of two to a power of a maximum number of bits to be searched, reading a value from a first level RAM and using the value to index to a second level RAM, and outputting, by the processor, a codeword length and a decoded codeword value from either the first or second level RAM.

公开(公告)号：US09112525B1

公开(公告)日：2015-08-18

申请号：US14672785

申请日：2015-03-30

Applicant: Comtech EF Data Corp.

Inventor： Seth Sjoholm ,  Edward Ray Coulter

IPC: H03M7/00 ,  H03M7/30 ,  G06F11/14 ,  H04L29/06 ,  G06F21/72 ,  H04L12/801

CPC classification number: H03M7/3084 ,  G06F11/1451 ,  G06F21/72 ,  H03M7/30 ,  H03M7/6023 ,  H04L29/06 ,  H04L47/10 ,  H04L69/04 ,  H04L69/22

Abstract: A method of parallel hash-based data compression comprising generating a hash value based on one or more original input data bytes of an incoming data stream using a processor, looking up, by the processor, a hash head comprising a first potential match location identified by the hash value and looking up, by the processor, at least one additional potential match location in a hash chain that comprises a linked list of potential match locations starting with a location indexed by the hash head. The method further comprises checking, by the processor, the at least one potential match location of the hash chain for one or more valid matches in a history buffer random access memory (RAM) and outputting, by the processor, a length and distance pair if a valid match is detected or the original input data byte if a match is undetected.

Abstract translation: 一种基于并行的基于散列的数据压缩的方法，包括：使用由处理器查找的处理器生成基于输入数据流的一个或多个原始输入数据字节的散列值的哈希值，该散列头包括由第一潜在匹配位置 所述散列值并由所述处理器查找散列链中至少一个附加的潜在匹配位置，其包括以由所述散列头索引的位置开始的潜在匹配位置的链接列表。 该方法还包括由处理器检查历史缓冲器随机存取存储器（RAM）中的一个或多个有效匹配的散列链的至少一个潜在匹配位置，并由处理器输出长度和距离对如果 检测到有效的匹配或未检测到匹配的原始输入数据字节。