IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0638265
(2009-12-15)
|
등록번호 |
US-8397120
(2013-03-12)
|
발명자
/ 주소 |
- Lor, Kar-Wing Edward
- Liew, Soung
- Wong, Vincent
|
출원인 / 주소 |
- Hong Kong Applied Science and Technology Research Institute Co. Ltd.
|
대리인 / 주소 |
Leydig, Voit & Mayer, Ltd.
|
인용정보 |
피인용 횟수 :
1 인용 특허 :
10 |
초록
▼
A method of error correction for a multicast message sent over a wireless network includes encoding a message into N data packets using a forward error correction code and multicasting at least L data packets of the N data packets over a wireless network to recipients. Recipients not receiving the a
A method of error correction for a multicast message sent over a wireless network includes encoding a message into N data packets using a forward error correction code and multicasting at least L data packets of the N data packets over a wireless network to recipients. Recipients not receiving the at least L data packets send a reply to the sender. The sender then selects a second subset of X data packets from the N data packets, and multicasts the X data packets over the wireless network to the recipients.
대표청구항
▼
1. A method of error correction for a multicast message sent over a wireless network, the method comprising: encoding a message into N data packets using a forward error correction code such that a subset of L data packets of the N data packets is needed to decode the message;multicasting at least K
1. A method of error correction for a multicast message sent over a wireless network, the method comprising: encoding a message into N data packets using a forward error correction code such that a subset of L data packets of the N data packets is needed to decode the message;multicasting at least K (where N>K≧L) data packets of the N data packets over the wireless network to a plurality of recipients;receiving a first reply from at least one recipient not receiving at least L data packets;selecting a second subset of X data packets from the N data packets by calculating a number J of data packets that the at least one recipient still needs to receive, where J=L−H and H is the number of data packets received among all recipients, andselecting X data packets from the N data packets such that X≧J, wherein X is selected such that X=J×R, where R is a redundancy factor and in a range from 1 to (N−L)/J, and X is in a range from J to (N−L); andmulticasting the X data packets over the wireless network to the plurality of recipients. 2. A method of error correction for a multicast message sent over a wireless network, the method comprising: encoding a message into N data packets using a forward error correction code, wherein the N data packets comprise at least L data packets needed to decode the message;selecting a first group of M data packets from the N data packets, wherein M is determined from L plus a number of redundant data packets related to a first packet loss probability, andM is in a range from L to N;multicasting the first group of M data packets over a wireless network to a plurality of recipients;receiving a first reply from at least one recipient, the reply comprising information on a number H of the first group of M data packets that have been multicast and received by the recipient;selecting a second group of X data packets from the N data packets using a second packet loss probability, wherein X is in a range from J to S, where J=L−H and is the number of data packets that the recipient still needs to receive to decode the message, and S=N−M; andmulticasting the second group of X data packets over the wireless network to the plurality of recipients, wherein M is the smaller of L/(1−E) and N, where E is an estimated unitized packet loss probability in a range from 0 to 1. 3. The method of claim 2 further including: receiving a second reply from at least one recipient that has not received the at least L data packets;selecting a subset of Xs data packets from the N data packets;multicasting the Xs data packets over the wireless network to the plurality of recipients; andrepeating the receiving, selecting, and multicasting until all recipients have received the L data packets or until all N data packets have been multicast to the recipients. 4. The method of claim 3 wherein all N data packets have been multicast to the recipients, and further including: receiving a further reply from the at least one recipient that has not received the at least L data packets; andunicasting missing data packets to the at least one recipient until the at least one recipient has received the L data packets. 5. The method of claim 2 further including: receiving a second reply from the at least one recipient that has not received the at least L data packets;selecting a subset of Xs data packets from the N data packets;multicasting the Xs data packets over the wireless network to the plurality of recipients;repeating the receiving, selecting, and multicasting until only one recipient has not received the L data packets; andunicasting missing data packets to the one recipient until the one recipient has received L data packets. 6. The method of claim 2 wherein the first reply from the at least one recipient includes a list of packets that the at least one recipient has received. 7. The method of claim 6 including selecting missing data packets from the N data packets multicast to the plurality of recipients and not received by the at least one recipient. 8. The method of claim 2 wherein the first reply from the at least one recipient uses Transmission Control Protocol. 9. The method of claim 2 including encoding a message into N data packets using a channel erasure coding method. 10. The method of claim 2 including multicasting the first group of M data packets at a first wireless transmission rate and multicasting the second group of X data packets at a second wireless transmission rate, wherein the first and second wireless transmission rates are different. 11. A method of error correction for a multicast message sent over a wireless network, the method comprising: encoding a message into N data packets using a forward error correction code, wherein the N data packets comprise at least L data packets needed to decode the message;selecting a first group of M data packets from the N data packets, wherein M is determined from L plus a number of redundant data packets related to a first packet loss probability, andM is in a range from L to N;multicasting the first group of M data packets over a wireless network to a plurality of recipients;receiving a first reply from at least one recipient, the reply comprising information on a number H of the first group of M data packets that have been multicast and received by the recipient;selecting a second group of X data packets from the N data packets using a second packet loss probability, wherein X is in a range from J to S, where J=L−H and is the number of data packets that the recipient still needs to receive to decode the message, and S=N−M;multicasting the second group of X data packets over the wireless network to the plurality of recipients; andcalculating a unitized packet loss probability P for the wireless network, where P=G/M and G is number of lost packets, M−H, and wherein selecting a second group of X data packets from the N data packets comprises selecting X data packets from the N data packets such that X=J×(1/(1−P)). 12. The method of claim 11 wherein receiving a first reply from the at least one recipient comprises receiving replies from a plurality of recipients,calculating P comprises calculating an average P for all recipients, anddetermining a J value for each of the recipients and J is the maximum of all of the J values determined. 13. The method of claim 11 including selecting a second group of X data packets from the N data packets such that X=min(S, Jmax×((1/(1−P)+r)), where r is a redundancy factor and at least 0, and Jmax is the maximum of all of the J values determined for all recipients. 14. A method of error correction for a multicast message sent over a wireless network, the method comprising: encoding a message into N data packets using a forward error correction code, wherein the N data packets comprise at least L data packets needed to decode the message;selecting a first group of M data packets from the N data packets, wherein M is determined from L plus a number of redundant data packets related to a first packet loss probability, andM is in a range from L to N;multicasting the first group of M data packets over a wireless network to a plurality of recipients;receiving a first reply from at least one recipient, the reply comprising information on a number H of the first group of M data packets that have been multicast and received by the recipient;selecting a second group of X data packets from the N data packets using a second packet loss probability, wherein X is in a range from J to S, where J=L−H and is the number of data packets that the recipient still needs to receive to decode the message, and S=N−M;multicasting the second group of X data packets over the wireless network to the plurality of recipients;receiving a second reply from at least one recipient that has not received the at least L data packets;selecting a subset of Xs data packets from the N data packets;multicasting the Xs data packets over the wireless network to the plurality of recipients;repeating the receiving, selecting, and multicasting until Ou(1−(B1/B2)). 16. The method of claim 15 where calculating P comprises calculating an average P for all recipients.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.