ERROR DETECTION METHODS IN
WIRELESS COMMUNICATION SYSTEMS
BACKGROUND OF THE INVENTION
1. Technical Field
The invention relates generally to wireless communication systems and, more particularly, to error detection methods for evaluating information that is transmitted in control channels of such systems. 10
2. Description of Related Art
Almost all wireless communication systems employ "frame-based" communication, where a certain number of bits, defined as a frame, are channel encoded together and transmitted. Most systems employ concatenated coding for 15 each frame with an inner error correction code such as a convolutional or Turbo error correction code and an outer error detection code.
FIG. 1 illustrates a process flow for forming a typical concatenated code structure at the transmitter of a base 20 station. As shown, an error detection code is added at 110 to a frame of a data channel, the frame comprising k bits. Typically a cyclic redundancy check (CRC) code, shown here as having a length p, is used as the error detection code. The CRC bits are computed based on the k information bits 25 . The CRC code is appended to the frame (e.g., k+p bits), and then passed through error correction encoding at 120. The error correction code is, for example, a convolutional code having a rate of 1/r, where r>l. After error correction encoding, the number of bits equals (r*(k+p)), which are 30 then passed to a modulator and transmitted over a channel. An error correction code decoder at the receiver of a user equipment (e.g., a mobile station) will attempt to correct any bit errors that take place on the channel.
In most cases, a frame is discarded when the receiver 35 detects an error, which is uncorrectable, in the transmission based on the error detection code. This results in a loss or delay of information, depending on whether a retransmission is subsequently carried out. The most widely used error detection code is the aforementioned CRC code. Standard 40 CRC codes include bit lengths of 8, 12, 16, 24 and/or 32 bits. The figure of merit or interest with error detection codes is the probability of an undetected error, i.e., a case where use of the inner error correction code could not correct transmission errors, and the outer error detection code did not 45 detect that the decoded information was erroneous. This is an undetected error because the decoded information is erroneous, but the error detection code did not catch the error. The undetected error probability with CRC codes is typically on the order of 2~L where L is the length of the 50 CRC. Thus, an 8-bit CRC has an undetected error probability of approximately 1/256.
The overhead associated with using a CRC is dependent on the number of information bits in the frame. Typically, the number of information bits of a frame such as frame k in 55 FIG. 1 exceeds several hundred bits, and thus any overhead effect of using an 8, 12 or 16-bit CRC is minimal. However, in certain applications very few bits need to be transmitted per frame and the overhead from using even a length-8 CRC may be excessive. One such example is in the transmission 60 of control channel information in wireless high speed data communication systems such as in the High Speed Downlink Packet Access (HSDPA) specification of the Universal Mobile Telecommunication System (UMTS) standard.
In HSDPA, transmission data for several user equipments 65 (hereinafter UEs, also frequently known as mobile stations) are multiplexed on a common high speed downlink shared
data channel (HS-DSCH). High data rates are obtained through scheduling, adaptive modulation and coding, and hybrid automatic repeat request (H-ARQ) as is known. UEs are scheduled on the shared data channel. The UEs are scheduled either in a purely time division multiplexed (TDM) manner, where all the available resources (power resources and data channelization codes) are assigned to one UE during a transmission time interval, or among multiple UEs in a transmission time interval (TTI). When transmitting to multiple UEs in a TTI, the power resources and data channelization codes are divided up among those UEs, not necessarily in a uniform manner. In the UMTS standard, the transmission time interval (TTI) is typically 2 ms or 3 timeslots (each timeslot being about 0.667 ms). Scheduling for the UEs is typically accomplished based on some type of information about the channel quality being experienced by the UE.
An important component of these high speed wireless systems is the use of a control channel. The control channel carries information related to (a) which UEs have been scheduled to receive a data transmission via a corresponding HS-DSCH (b) what data channel codes, are assigned to each particular UE, and (c) modulation and HARQ-related information. From a system efficiency perspective, a few control channels are defined such that they are shared among all UEs, rather than providing a dedicated control channel per UE.
An exemplary configuration is to define up to M high speed shared control channels (HS-SCCHs) for simultaneous transmissions, where M=4, for example. For each TTI, each HS-SCCH carries HS-DSCH-related downlink signaling for one UE. The number of HS-SCCHs may range from a minimum of one HS-SCCH (M=l) to a maximum of four HS-SCCH's (M=4). This is the number of HS-SCCH's as seen from the UE's point-of-view. In other words, a UE determines whether an ensuing transmission on any of the HS-DSCHs is intended for itself or not only upon or after decoding information in the HS-SCCHs.
FIG. 2 illustrates the relationship between HS-SCCHs 210 and their corresponding shared HS-DSCH counterparts 220. In FIG. 2, each HS-SCCHx (x=l to 4) carries information pertinent to a corresponding HS-DSCHx (x=l to 4). The number of HS-DSCHs, and therefore the number of HS-SCCHs that may be used, can vary for each TTI, depending on the number of UEs being simultaneously scheduled in the TTI. Accordingly, the configuration of HS-SCCHs and HS-DSCHs in FIG. 2 enables the data channelization codes and power resources to be divided among four simultaneous transmissions.
Referring again to FIG. 2, control channel data on each HS-SCCH is typically divided into two parts. Part I consists of information relating to those data channelization codes that have been assigned to a particular UE, for example. Part II data contains HARQ related information, and other transport information. To maintain complexity low at the UE, HS-SCCH designs typically allow Part I information to be transmitted prior to the commencement (i.e., before t=0) of data transmission, as shown in FIG. 2. Accordingly, with the current configuration, each UE must decode Part I on every HS-SCCH, in every TTI, in order to determine (a) whether or not the transmission was intended for that particular UE, and (b) if the transmission was intended for that particular UE, the UE must decode Part I and figure out what channelization codes the corresponding HS-DSCH will arrive on.
Therefore, each UE must decode up to four (4) HSSCCHs in every TTI, prior to commencement of data transmission. From a UE processing complexity perspec