1. Technical Field
The invention relates to reducing the effect of Doppler and/or high frequency errors in a cellular system.
2. Discussion of Related Art
This invention arose in the context of developments underway in the UL part of UTRAN long term evolution (LTE) often referred as 3.9G but is not limited to that context.
In 3G LTE a scheduling request channel for the uplink is in the process of being defined. The current discussion and decision is that the channel will utilize a code domain spreading and multiplexing between users. The performance of the channel in a multi-access scenario has been presented in 3GPP document R1-071663. The used scrambling codes are proposed to be Zadoff-Chu sequences. The basic idea of the code domain spreading and multiplexing is shown in
A block diagram of a transmitter for the scheduling request channel is shown in
The random access preambles are generated from Zadoff-Chu sequences with zero correlation zone, ZC-ZCZ, generated from one or several root Zadoff-Chu sequences. The network configures the set of preamble sequences the UE is allowed to use.
The uth root Zadoff-Chu sequence is defined by
where the length NZC of the Zadoff-Chu sequence is given by Table 14. From the uth root Zadoff-Chu sequence, random access preambles with zero correlation zone are defined by cyclic shifts of multiples of NCS according to
x
u,v(n)=xu((n+vNCS)mod NZC)
where NCS iS given by Table 14.
The multiplexing between the different users is achieved through the code domain orthogonality. Cyclic shifts of Zadoff-Chu sequences are used as the orthogonal codes. The maximum number of orthogonal codes can be computed as above for the resource elements as 12*7=84. The orthogonality within a single block, or FDMA symbol, is limited by the channel delay spread and the sinc pulse shape used in the transceiver. Between the blocks the orthogonality is limited by the channel Doppler spread as well as the frequency error. In practice, the number of orthogonal codes can be less than 84 due to these phenomena.
In this patent application, we refer to the different orthogonal codes by an index. An orthogonal code is a combination of spreading within the block and between the blocks. The index is determined by placing the different orthogonal codes in a vector and incrementing first the cyclic shifts between the blocks. An example for 3 cyclic shifts within a block and 7 cyclic shifts between blocks is given in
For the cyclic shifts between the blocks, it is possible to select from six (1-6) different root Zadoff-Chu sequences. For the cyclic shifts within the blocks, a selection from four (1, 5, 7, 11) root sequences is made.
Current solutions do not show the performance of the code domain multiplexing scheme in a high speed environment, nor in environments with a high speed line-of-sight channel condition. These channel conditions cause a large degradation in performance. The performance falls to an unsatisfactory level, thus needing some corrective measures.
The invention is based on the observation that high Doppler frequencies as well as high frequency errors will cause a loss in the orthogonality between the otherwise orthogonal codes. The loss is due to side peaks appearing in the auto-correlation function. A regularity can be observed in the side peaks and the index of the codes which are most interfered by the side peaks.
Accounting for both positive and negative frequency errors is needed. The index is dependant on the root sequence used for spreading between the blocks as well as the actual index of the orthogonal code. These side peaks begin to “look like” other orthogonal codes, causing a sharp increase in false alarms.
Motivated by these observations, a method for reducing the set of orthogonal codes to mitigate the issue is proposed. For the different root sequences used for spreading between blocks, the main side peaks with reference to the actual orthogonal code index are listed.
For root sequence 1, the main side peaks are formed to indexes −1, +1, −6 and +6 relative to the used orthogonal code index. The actual index is determined as a modulus(L)+1 operation, where L is the total number of orthogonal codes being used.
For root sequence 2, the main side peaks are formed to indexes −4, +4, −3 and +3.
For root sequence 3, the main side peaks are formed to indexes −5, +5, −2 and +2.
For root sequence 4, the same rule as root sequence 3.
For root sequence 5, the same rule as root sequence 2.
For root sequence 6, the same rule as root sequence 1.
The invention is characterized by the following features:
It is to be understood that all presented exemplary embodiments may also be used in any suitable combination.
Other objects and features of the present invention will become apparent from the following detailed description considered in conjunction with the accompanying drawings. It is to be understood, however, that the drawings are designed solely for purposes of illustration and not as a definition of the limits of the invention, for which reference should be made to the appended claims. It should be further understood that the drawings are not drawn to scale and that they are merely intended to conceptually illustrate the structures and procedures described herein.
The transmitter in
The scheduling request channel utilizes code domain multiplexing of different transmitters. Different cyclic shifts of the same CAZAC root sequences are typically used within the same cell, while different root sequences are typically used in different cells.
We give an identifier for the orthogonal code used by the transmitter. First we indicate the root sequence used for sequence 1 and sequence 2. Then we identify the cyclic shift used by sequence 1 and sequence 2. We give a value 1 for no shift, 2 for a shift by 1, etc. We note that a shift by the length of the sequence is equal to no shift.
Referring to
Without some special consideration, the performance of the transmitter scheme in a high Doppler channel or a channel with a high frequency error is poor. The invention provides a mechanism for improving the performance significantly in these conditions.
All sets of length 9, which fulfil the criteria, for L=21 is given (where L is the total number of orthogonal codes being used). Nine is the maximum number of sequences, which can be found for L=21. A similar table can be found for different values of L. The root sequence refers to the root sequence used for spreading between the blocks.
Root sequence 4: (same as Root sequence 3)
Root sequence 5: (same as Root sequence 2)
Root sequence 6: (same as Root sequence 1)
The signal processor of the user equipment may take the form shown in
The signal processor of the base station may also take the form shown in
This application claims priority from U.S. Provisional Application Ser. Number 60/955,926 filed Aug. 15, 2007.
Number | Date | Country | |
---|---|---|---|
60955926 | Aug 2007 | US |