Error detection/correction and fault detection/recovery – Pulse or data error handling – Digital data error correction
Reexamination Certificate
1999-09-17
2003-04-22
Decady, Albert (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Digital data error correction
Reexamination Certificate
active
06553540
ABSTRACT:
FIELD OF THE INVENTION
The present invention is related to telecommunications systems, such as wireless local area networks. More particularly, the present invention relates to the protection of information bits, being transmitted within a telecommunications system.
BACKGROUND
In response to a growing need for low-cost, short-range, high capacity radio links, the European Telecommunications Standards Institute (ETSI) has established a standardization project for Broadband Radio Access Networks (BRAN). One of the broadband radio access networks being developed under ETSI BRAN is HIPERLAN Type 2 (HIPERLAN/2). HIPERLAN/2 is a short-range, high data-rate system that offers high speed access (i.e., up to 54 Mbit/sec) to a variety of networks including Universal Mobile Telecommunications System (UMTS) core networks, Asynchronous Transfer Mode (ATM) networks and Internet Protocol (IP) based networks.
An important feature of HIPERLAN/2 will be the centralized medium access control (MAC) protocol, which is intended to provide an efficient use of the available spectrum. In accordance with the MAC protocol, an access point (AP), also referred to as a base station, controls channel access by assigning downlink and uplink timeslots to the various mobile terminals (MTs) with which it is communicating, wherein a MT receives data from the access point during a downlink timeslot and transmits data to the access point during an uplink timeslot.
A characteristic of HIPERLAN/2 is that the data is transported by protocol data units (PDUs). There are different PDU types. For instance, for transporting control information, there are control PDUs, and for transporting actual data, there are data PDUs, wherein each PDU of a certain type has a fixed size.
The HIPERLAN/2 standard specifies three system layers, namely, the physical layer, the data link control (DLC) layer comprising logical link control (LLC) and MAC, and the convergence layer (CL). The CL is the interface between higher layers and the DLC layer. For example, there may be a CL for TCP/IP, which segments the IP packets into data PDUs. The DLC layer adds header information before the PDUs are passed to the physical layer. While the following discussion focuses on data PDUs, it will be understood that the same applies to control PDUs.
The physical layer of HIPERLAN/2 will be based on orthogonal frequency division multiplexing (OFDM) and convolutional encoding. The granularity of data units on the physical layer is therefore an OFDM symbol. Depending on the subcarrier modulation scheme, e.g. BPSK, QPSK, 8PSK, 16QAM or 64QAM, the number of OFDM symbols needed to carry one PDU will be different.
Another feature of HIPERLAN/2 is that several physical layer modes will be provided. For example, the system may provide physical layer modes based on the aforementioned modulation schemes and convolutional codes for rates of ½, {fraction (9/16)} and ¾. An important requirement for the DLC design is that the physical layer modes shall be designed such that each PDU fits into an integer number of OFDM symbols. Otherwise, capacity is wasted by using e.g. bit padding.
As an example, where 48 subcarriers are used for data and each PDU comprises 54 bytes, a BPSK modulation scheme with a code rate ½ is used. In this case, there are 48 bits carried by each symbol. Using a code rate of ½ for the 432 input bits (i.e., 54 bytes*8 bits/byte=432 bits) results in
864
encoder output bits without tail bits. These 864 encoder output bits are carried by exactly 18 OFDM symbols (i.e., 864 bits/(48 bits/symbol)=18 symbols). Because there is an integer number of OFDM symbols, no padding bits are needed. This is true for all other modes in HIPERLAN/2, so long as the tail bits are discarded.
It should be highlighted that the code rates ½, {fraction (9/16)} and ¾, which are being discussed for HIPERLAN/2 are only precise when the tail bits are disregarded. This issue will be discussed further below.
Another WLAN system currently being standardized is the IEEE802.11 system. The IEEE802.11 system is being designed with a 5 GHz mode, which will have similar physical layer parameters with respect to HIPERLAN/2. However, the IEEE802.11 system is specifically being designed for transmitting IP packets by radio, where the protocol principles are similar to Ethernet; hence, the MAC protocol will be very different from HIPERLAN/2. In an IEEE802.11 system, for instance, IP packets, or segments thereof, having variable lengths are transmitted. The code rates which are currently being considered for IEEE802.11 are ½, ⅔ and ¾.
An example of a flexible MAC frame
100
of HIPERLAN/2 is depicted in FIG.
1
. As shown, the MAC frame
100
includes a broadcast control channel (BCCH), which contains information that is transmitted over the whole area (e.g., cell) covered by one AP. The assignment of logical channels to different MTs is transmitted in the frame control channel (FCCH), sometimes referred to as the resource grant channel. Accordingly, each MT knows the exact, dedicated time period in the MAC frame
100
when it is expected to receive a downlink burst and/or send an uplink burst. A random access channels (RACH) is located at the end of the MAC frame
100
. A MT may request capacity by transmitting the request in its assigned uplink burst channel or via the random access channel.
The described MAC frame
100
illustrated in
FIG. 1
should be understood as one possible arrangement of fields. In fact, the fields may appear in a different order. Furthermore, some of the fields in the MAC frame
100
may not appear at all, while others may be added. Regardless, the present invention described below is still applicable.
In each MAC frame field, data is transmitted from the AP to one or more MTs, or vice versa. A block of data which is destined for, or sent by, one MT is called a “burst”. Each burst comprises one or more PDUs. On the DLC layer, the concatenation of several PDUs may also be called a PDU train, or ‘cell’ train when the transmission of ATM cells is involved. On the physical layer, a preamble may be added at the beginning of each burst for synchronization and channel estimation purposes. If the channel access scheme is dynamic TDMA, the length of a burst is variable.
A convolutional code (CC) can be used to encode blocks of data. When CCs are used, tail bits (e.g. zero bits) are appended to the stream of information bits. The tail bits ensure that the encoding process terminates in a pre-defined state, e.g. in the zero state, thus providing protection for the last bits in a block. For a CC with constraint length 7, 6 tail bits are needed for termination. This results in additional redundancy. However, the code rate of a CC is often given without taking into account the tail bits. For example, the code rates ½, {fraction (9/16)} and ¾, which are being discussed for HIPERLAN/2, do not include the tail bits. Therefore, the actual code rate is slightly lower due to the redundancy being increased as caused by the use of additional tail bits.
In a TDMA system with fixed timeslots, e.g., GSM, the timeslots have a fixed duration, and while the number of information bits may vary, the number of modulating bits remains fixed. This is often provided by a variety of puncturing schemes, one per physical layer mode. The tail bits are included in the design of the puncturing scheme which is specific.
IEEE802.11 is an ad-hoc network without regular frame structure. IP packets or segments thereof, which have variable length or more precisely a length being typically determined by the IP layer above the IEEE802.11 protocol, are transmitted. The encoding is performed based on a selected coding scheme with a code rate for the whole packet. At the end of the packet, the tail bits are appended and encoded like the data. The encoded data, including the tail bits, is mapped to OFDM symbols. The last OFDM symbol may not be filled totally, hence bit padding is applied.
HIPERLAN/2, and like systems, are char
Schramm Peter
Wachsmann Udo
De'cady Albert
Lamarre Guy
Telefonaktiebolaget LM Ericsson
LandOfFree
Efficient system and method for forward error correction does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Efficient system and method for forward error correction, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Efficient system and method for forward error correction will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3025251