Dynamic wait acknowledge for network protocol

Details Dynamic wait acknowledge for network protocol Dynamic wait acknowledge for network protocol

1999-03-17

2003-10-14

Yao, Kwang Bin (Department: 2662)

Multiplex communications

Pathfinding or routing

Switching a message which includes an address header

C370S389000, C370S236000

Reexamination Certificate

active

06633574

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to network protocols, and particularly to acknowledgement of receipt of messages in the network protocols.
BACKGROUND OF THE INVENTION
Protocols for control networks, such as the internet, are well known. One such protocol is the protocol developed by ECHELON SYSTEMS CORPORATION. Its protocol is described in U.S. Pat. No. 5,018,138. Applicant hereby incorporates this U.S. Patent by reference.
One of the basic precepts of an internet protocol is the notion of an acknowledgment service.
FIG. 1
illustrates a portion of a conventional network. The network
100
comprises a plurality of subnets
102
a
-
102
c
coupled to a plurality of nodes
104
a
-
104
f
. Routers
106
a
-
106
b
are located between subnets
102
a
-
102
c
. The routers
106
a
-
106
b
route messages to the appropriate subnet
102
a
-
102
c
. When an originating node, such as node
104
a
, sends a message to a destination node, such as node
104
f
, the message travels from subnet
102
a
, through router
106
a
, to subnet
102
b
, through router
106
b
, to subnet
102
c
, and then to destination node
104
f
. The sequence of routers
106
a
-
106
b
through which the message travels is the path of the message from the originating node
104
a
to the destination node
104
f.
FIG. 2
is a state diagram illustrating a conventional acknowledgement protocol. Referencing now to
FIGS. 1 and 2
together, before a message is sent, an originating node
104
a
has an idle state
202
. When the originating node
104
a
wishes to send a packet or message, it has a contention state
204
. Once the message is sent, the originating node
104
a
initiates or sets a timer indicated by a state of waiting for an acknowledgment packet
206
from the destination node
104
f
. The time period of length of the timer is based upon assumptions concerning how long it takes to transport a message through the various elements of the network and to get a message back. This time period is dependent upon the physical media itself and how long it takes for a message to travel along this media. It is advantageous to minimize the time period as an overly long time period wastes valuable network resources. A predetermined amount of time is given for each portion of the route from a subnet to a router, called a “hop”. Once destination node
104
f
receives the message, it sends an acknowledgement packet back to the originating node
104
a
. If the originating node
104
a
receives the acknowledgement packet before the timer expires, then it knows that the delivery of the message was successful. The originating node
104
a
returns to an idle state
202
. If the destination node
104
f
never receives the message due to problems somewhere along the message's path, then the originating node's
104
a
timer will expire without receiving an acknowledgement packet. In this situation, originating node
104
a
assumes that the delivery of the message failed. It may then retry message delivery by resending the message. The protocol typically allows for a predetermined number of retries before abandoning the attempt to deliver the message. This protocol is well known in the art and will not be discussed in further detail here.
One problem which exist for this conventional protocol is that the amount of time it take to transport a message through each hop is actually dynamic and variable. The distance in each hop are variable. Some may be very long while others are very short. A message may travel through the routers at different rates, depending on the characteristics of the individual router. When the actual time for a message to be delivered is actually very long, a message may be received by a destination node but its acknowledgement packet may not be able to reach the originating node before the originating node's timer expires. The originating node would be forced to resend the message, wasting valuable network resources.
One effect of this problem is when the originating node retries delivery of the message, the retry message may be mistakenly considered a new message under certain circumstances. For example, although the destination node can detect if duplicate message arrive, if enough time lapses, or new messages are received from other nodes, before the retry message is received, the retry message may be interpreted as a new message by the destination node. This could cause undesirable results. For example, assume that the opening and closing of doors in a water dam is accomplished through a command which toggles the state of the doors. Assume also an operator sends a command to close the water dam doors. However, this message delivery fails. The originating node then retries delivery by resending the message. Due to the problem described above, the second message is mistakenly perceived as a new message. The first message toggles the dam doors closed, while the second message toggles the dam door open again, resulting in the opposite of the desired result. In other applications, the user may receive strange behavior from the network without knowing why, since the problem is difficult, if not impossible, to trace.
Accordingly, what is needed is a method for a dynamic wait acknowledge for a network protocol. The present invention addresses this need.
SUMMARY OF THE INVENTION
The present invention is a method for acknowledgment of receipt of a message in a network, where the network includes a plurality of nodes. The method includes sending a message from a first node of the plurality of nodes destined for a second node of the plurality of nodes; setting a timer for the first node for receipt of an acknowledgment packet from the second node; determining if a wait acknowledge packet is received by the first node; and resetting the timer if the wait acknowledge packet is received by the first node. The present invention adds a wait acknowledge event to the state in the acknowledgment protocol which waits for an acknowledgement packet from the destination node. The wait acknowledge packet notifies the recipient that the message sent by the originating node has been received and forwarded. As long as the message is being properly forwarded to the next hop, wait acknowledge packets will continue to be sent to the originating node. In this manner, the amount of time the originating node waits for the acknowledgment packet from the destination node is dynamic and is dependent upon the actual length of time it takes for the message to be transmitted through the network. This method reduces waste of network resources and minimizes the problem of unintentional duplicate messages.


REFERENCES:
patent: 4941143 (1990-07-01), Twitty et al.
patent: 4947484 (1990-08-01), Twitty et al.
patent: 4955018 (1990-09-01), Twitty et al.
patent: 4969146 (1990-11-01), Twitty et al.
patent: 4969147 (1990-11-01), Markkula, Jr. et al.
patent: 4970714 (1990-11-01), Chen et al.
patent: 4975904 (1990-12-01), Mann et al.
patent: 5297143 (1994-03-01), Fridrich et al.
patent: 5434860 (1995-07-01), Riddle
patent: 5519699 (1996-05-01), Ohsawa
patent: 5612957 (1997-03-01), Gregerson et al.
patent: 5854899 (1998-12-01), Callon et al.
patent: 5914936 (1999-06-01), Hatono et al.
patent: 5940371 (1999-08-01), Mitts et al.

Affiliated with

Also associated with

Rating

Dynamic wait acknowledge for network protocol does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Dynamic wait acknowledge for network protocol, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dynamic wait acknowledge for network protocol will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3156314