Wireless broadcast protocol

• US 8,091,005 B2
• Filed: 11/20/2008
• Issued: 01/03/2012
• Est. Priority Date: 11/14/2003
• Status: Expired due to Fees

• Current Assignee: Avocent Huntsville Corporation (Vertiv Holdings Co.)
• Original Assignee: Avocent Huntsville Corporation (Vertiv Holdings Co.)
• Inventors: Huntly-Playle, Iain, Covington, C. David, Lally, Charles, Li, Sheng, Moore, Richard L.
• Primary Examiner(s): Torres, Joseph D
• Application Number:
  US12/292,541
• Publication Number:
  US 20090083436A1
• Time in Patent Office: 1,139 Days
• Field of Search: 714/750, 714/748, 714/749
• US Class Current: 714/750
• CPC Class Codes: G06F 11/1443 Transmit or communication errors

Abstract

In a keyboard-video-mouse (“KVM”) system in which a target computer may be wirelessly accessed by a plurality of remote stations, a method includes, by the target computer: obtaining a frame of video data; transmitting packets for the frame; transmitting a query packet; obtaining a list of requests from at least one of the remote stations, each request from a particular remote station identifying packets missed by that particular remote station; and resending at least some of the requested packets.

Claims

1. In a keyboard-video-mouse (“

KVM”

) system in which a target computer may be wirelessly accessed by a plurality of remote stations, a method comprising the steps of,

(A) by the target computer;

(a1) obtaining a frame of video data;

(a2) compressing in a video compressor at least some of the video data using a line fitting compression scheme to produce at least one sequence of absolute coded pixel segments and relative coded pixel segments which combine to represent the video data;

(a3) transmitting packets for the frame;

(a4) transmitting a query packet requesting at least some of said plurality of remote stations to provide a list of packets that need to be resent;

(a5) obtaining from each of at least two of the remote stations, a list of one or more packets that need to be resent, wherein the list of packets from a particular remote station identifies packets missed by that particular remote station;

(a6) removing duplicates from the lists of packets that need to be resent; and

then (a7) resending at least some of the requested packets; and

(B) by each of a plurality of the remote stations;

(b1) receiving from the target computer broadcasted packets associated with the frame of video data;

(b2) keeping a list of missed packets;

(b3) upon receipt of the query packet from the target computer, sending to the target computer at least some entries from the list of missed packets; and

(b4) receiving at least some resent packets rebroadcasted from the target computer including some entries from the list of missed packets and other non-duplicated entries from corresponding lists of missed packets from others of said plurality of the remote stations.

2. A method as in claim 1 wherein the list has no more than eight entries.

3. The method as in claim 1 further comprising:

the target computer repeating steps (a1) through (a7) while there is more video data.

4. The method as in claim 1 further comprising:

the target computer sorting the lists of packets that need to be resent prior to removing duplicates.

5. The method of claim 1 further comprising:

the target computer compressing in the video compressor at least some of the video data using a line fitting compression scheme to produce at least one sequence of absolute coded pixel segments and relative coded pixel segments which combine to represent the video data;

wherein the video data represent red, green and blue (RGB) pixel components and wherein a segment is formed to be as long as possible such that the curvature of each of the red, green, and blue component values that form the segment is constrained to be less than a threshold, and wherein the measure of curvature is the magnitude of the output of a finite impulse response filter with coefficients (1, −

2,

     1) applied to the RGB pixel component values forming each segment, and wherein the curvature is measured as the magnitude of the output of each of a set of finite impulse response filters with coefficients (1, −

2,

     1) and other filter coefficients derived from said finite impulse response filter by interspersing zeros, specifically producing first (1, 0, −

2, 0,

     1), then (1, 0, 0, −

2, 0, 0,

     1), and so on up to a maximum segment length.

6. The method as in claim 5, wherein each absolute coded pixel segment represents a single pixel and encodes the most significant bits of each color component of that pixel.

7. A method as in claim 6 wherein the most significant bits of each color component of a pixel comprise the top five bits of each color component of that pixel.

8. A method as in claim 7, wherein each relative coded pixel segment represents a sequence of pixels and defines a difference in value between a starting point of the first pixel in the sequence and an ending point of the last pixel in the sequence.

9. A method as in claim 8 wherein the sequence of pixels is between two and nine pixels in length and wherein the relative coded pixel segment provides a signed delta value of the difference between the starting point and the ending point of the segment.