Communication device and method

Details Communication device and method Communication device and method

1999-10-20

2003-03-11

Vincent, David (Department: 2666)

Multiplex communications

Data flow congestion prevention or control

Control of data admission to the network

C370S235000, C370S469000, C370S474000, C714S001000

Reexamination Certificate

active

06532211

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to a communication device and method for data unit based communication, where implementations of at least a first and second communication protocol are used, and PDUs (Protocol Data Units) of a third, upper layer protocol are segmented into PDUs of the second, lower layer protocol, and these lower layer PDUs are sent over a physical connection in accordance with the first protocol, which provides adjustable reliability levels for the lower layer PDUs.
BACKGROUND OF THE INVENTION
As is known in the art of communication, protocols are sets of rules with which two points can exchange data units in a defined way. Two implementations of a protocol at two points that exchange data units are also referred to as peers. For the purpose of the present specification, the term data unit or protocol data unit (PDU) will refer to the finite data carrier specified by a given protocol. It may be noted that with respect to different protocols, different terms are used for the PDUs. For example, the data units of the internet protocol (IP) are referred to as packets, whereas the data units of the point-to-point protocol (PPP) are referred to as frames. All such terms, i.e. frames, packets etc. fall under the general term data unit.
Furthermore, the concept of layering different protocols is also well-known. According to this concept, data units of one protocol are embedded into data units of another protocol when being sent, and are extracted when being received. The term “embedding” refers to both the possibility of encapsulation as well as segmentation.
FIG. 2
shows a generic stack, and the figure introduces a number of terms that will be used as examples and for explanatory purposes in the following description. The stack shown in
FIG. 2
shows five layers. Naturally, the number of layers can be larger. L
3
refers to a network layer protocol, e.g. the internet protocol IP. L
4
refers to a protocol above the network layer, e.g. the transmission control protocol TCP. L
4
is also to be seen as representing all protocols that may lie above. L
2
_Frame refers to a link layer protocol which embeds or frames L
3
PDUS, for example the point to point protocol PPP, which is typically used for circuit-switched data in systems operating in accordance with the GSM (global system for mobile communication) standard. Other examples would be LLC (logical link control; defined in standard GSM 04.64) used for GPRS (General Packet Radio Service; defined in standard GSM 03.64) or W-CDMA (wide band code division multiple access). L
2
_ARQ refers to a link layer protocol that can segment L
2
_Frame PDUs into smaller L
2
_ARQ PDUs and implements an automatic repeat request function ARQ on the basis of these L
2
_ARQ PDUS. Automatic repeat request (ARQ) means that the protocol supports an automatic retransmission of PDUs under predetermined conditions. Examples of an L
2
_ARQ protocol are the radio link protocols RLP used for circuit-switched data in GSM, the radio link control protocol (RLC) used for GPRS and the RLCP (Radio Link Control Protocol) used for W-CDMA.
L
1
refers to a physical layer protocol or a combination of physical layer protocols that can operate on the basis of single or plural L
2
_ARQ PDUs. The L
1
protocol is to be understood as a protocol that can provide at least two different reliability levels for the transmission of L
2
_ARQ PDUs. Examples of the L
1
protocol are FEC protocols (forward error control) or power control protocols, or a combination of both. For example, different L
2
_ARQ PDUs can be protected with varying strength of forward error control and/or with varying transmission power. Further possibilities for adjusting the transmission reliability, which may be used individually or combined, are changing the spreading factor in CDMA or W-CDMA, the interleaving depth, the modulation or the antenna diversity. As these concepts are known in the art, a further description is not necessary.
It may be noted that the L
2
_Frame protocol is optional, as it would also be possible that the L
3
protocol PDUs are directly segmented by the L
2
_ARQ protocol, without first being encapsulated into L
2
_Frame protocol data units.
Commonly, the L
1
protocol will have a general adoption mechanism for deciding on the reliability level that is to be set for each L
2
_ARQ PDU. Different known physical layer protocols provide different adoption mechanisms, e.g. the setting of the reliability level may depend on the quality of the physical link over which data units are being sent.
Such an arrangement can lead to a number of problems. Many L
3
protocols and protocols running on top of L
3
are sensitive to a delay per data unit and can perform badly or even fail if the delay per data unit exceeds certain bounds. The problem is that when the L
3
protocol is running over L
2
_Frame/L
2
_ARQ or on L
2
_ARQ directly, the L
2
_ARQ protocol can introduce an additional delay per L
3
PDU, due to the retransmission of L
2
_ARQ PDUs. This additional delay is basically unbounded and can cause considerable problems. This will be explained in connection with the diagram shown in FIG.
3
.
For the following explanation, it will be assumed that L
2
_Frame PDUs are being segmented by the L
2
_ARQ protocol, but as already mentioned above, it is equally well possible that L
3
protocol PDUs are directly being segmented.
FIG. 3
a
shows one L
2
_Frame PDU, and this higher layer PDU is segmented into a given number of L
2
_ARQ data units. Two of these L
2
_ARQ data units are marked as a and b, respectively, for the purpose of a later explanation. As also indicated in
FIG. 3
a
, the L
2
_Frame PDU has a given transmission delay, i.e. a given length, just as the L
2
-ARQ data unit has a given length or transmission delay.
As show n in
FIG. 3
b
, the following problem can occur. If the L
2
_ARQ data unit a has to b e re transmitted, then the number of L
2
_ARQ data units that needs to be sent is increased by one, and correspondingly the transmission of the L
2
_Frame PDU is delayed by the transmission delay of one L
2
-ARQ data unit. However, if the L
2
_ARQ data unit marked as b, which lies at the end of the L
2
_Frame PDU, has to be retransmitted, then this will delay the transmission of the L
2
_Frame PDU by the round trip time (RTT) of the L
2
_ARQ layer. The round trip time RTT of a layer is basically the time that passes between the sending of a data unit of that layer by a sending peer, and the receipt by the sending peer of the acknowledgment message that confirms that the given data unit was received at the other end by the receiving peer. The L
2
_ARQ RTT is typically much longer than the L
2
_ARQ Transmission delay.
As already mentioned, this delay can cause significant problems in higher layers.
Another problem can occur in systems where the L
2
_ARQ peer uses window-based flow control. Window-based flow control is well-known in the art and basically means that the flow control is accomplished in accordance with a defined number of consecutive octets or bits that is referred to as a send window, where the allowed number of unacknowledged data units is limited to said send window. In other words, flow control is such that in a given series of data units to be sent, a certain number of data units following a given data unit may be sent out, even though the safe receipt of said given data unit has not yet been acknowledged, but this number of unacknowledged data units is limited to the send window. As already mentioned, this concept is well-known in the art, see e.g. TCP/IP Illustrated, Vol. 1, The Protocols, by W. Richard Stevens, Addison-Wesley Longman, Inc. 1994. A further explanation is therefore not necessary.
In an L
2
_ARQ peer that uses window-based flow control, the sender cannot send any new L
2
_ARQ PDUs when the send buffer is exhausted with back-logged copies of L
2
_ARQ PDUs that have already been sent but not acknowledged by the receiving L
2
_ARQ peer. This will briefly be explained in connection with FIG.
9
. This figure shows a c

Affiliated with

Also associated with

Rating

Communication device and method does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Communication device and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Communication device and method will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3054936