RU2251216C2 - Method for finding a cell in cdma communications system - Google Patents

Abstract

FIELD: mobile communications.

SUBSTANCE: method includes setting a codes set having certain properties, consisting of Q-numbered code words having length M, including symbols from set of Q short codes. First property is, that none cyclic displacement of code word produces correct code word as a result. Other properties are presence of mutually unambiguous match between long code message and correct code word. In case of noise and interference decoder, at acceptable complication levels, provides for search of both random displacement (in such a way determining frame synchronizations) and transmitted code word (i.e. long code indication message related thereto).

EFFECT: higher efficiency.

5 cl, 22 dwg, 5 tbl

Description

FIELD OF THE INVENTION

The present invention relates generally to spread spectrum data transmission systems, and more particularly, to a cell search process performed by a mobile station to provide synchronization with a base station and to retrieve a long code of a specific cell, as well as frame synchronization information used in a multiple access communication system. code division multiplexing (CDMA).

BACKGROUND

The cell phone industry has achieved phenomenal success in commercial operations worldwide. Growth in the main metropolitan areas far exceeded expectations and ahead of the bandwidth of the communication system. If this trend continues, the effect of rapid growth will soon reach even the smallest markets. The dominant problem associated with continuous growth is that the consumer base is expanding, while the size of the electromagnetic spectrum allocated to cellular service providers for use in communication networks transmitting on radio frequencies remains limited. Creative solutions are needed to meet this increasing bandwidth requirement with a limited range of available services, as well as to maintain high-quality service and to prevent price increases.

Currently, when accessing a channel, frequency division multiple access (FDMA) and time division multiple access (TDMA) methods are primarily used. In FDMA communication systems, a physical channel contains one radio frequency band in which signal transmission power is concentrated. In TDMA communication systems, the physical communication channel contains a time interval in a periodic series of time intervals at the same radio frequency. Despite the fact that when using the FDMA and TDMA communication systems, acceptable efficiency is achieved, usually due to increasing customer requests, the communication channel is overloaded. Accordingly, alternative methods of access to the communication channel are currently being proposed, considered and implemented.

An extended spectrum is a data transmission method that finds commercial application as a new method of accessing a channel in wireless communication systems. Spread spectrum data transmission systems have appeared since the Second World War. The early applications were mostly military-oriented (with respect to providing effective interference, radar and satellites). However, there is an increasing interest in the use of spread spectrum data communication systems in other communications systems applications, including digital cellular radio, terrestrial mobile radio and internal / external personal communications networks.

The spread spectrum data transmission systems operate in a completely different way than the standard TDMA and FDMA communication systems. For example, in a spread spectrum transmitter with a direct sequence code in a CDMA (CDMA-PP) system, the stream of digital symbols for a given specialized or common channel having a basic symbol rate is expanded at the elementary signal rate. This expansion operation involves applying a unique channel extension code (sometimes called a signature sequence) to a character stream that, when redundancy is added, increases its transmission speed (bandwidth). Typically, the stream of digital characters when expanded is multiplied by a unique digital code. An intermediate signal containing the resulting data sequence (elementary signals) is then added to other, similarly processed (i.e., extended), intermediate signals related to other channels. Further, to generate an output signal for multi-channel data transmission via a communication medium, a scrambling code unique to the base station is applied to the summed intermediate signals (often called the “long code”, since it is, in most cases, longer than the spreading code). After that, intermediate signals related to specialized / common communication channels successfully share the same frequency of communication tools for data transmission, with many signals placed on top of each other, both in the frequency domain and in the time domain. Since the spreading codes used are unique for the channel, each intermediate signal transmitted through the common frequency of the communication means is also unique, and by applying appropriate processing methods to the receiver can be extracted from other signals.

At a spread spectrum receiver of a CDMA-PP system mobile station, the received signals are reconstructed using, i.e., by multiplying or matching the corresponding scrambling and spreading codes to compress or remove the encoding from the useful transmitted signal and return to the main symbol rate. However, where the spreading code is used for other transmitted and received intermediate signals, the result is only noise. The compression operation, therefore, does include a correlation process that compares the received signal with the corresponding digital code to recover useful information from the communication channel.

Before a connection can occur between the base station and the mobile station of the spread spectrum data system, or information can be transmitted on the radio frequency, the mobile station must find and synchronize itself with the synchronization reference signal of this base station. This process is commonly referred to as cell search. In a CDMA-PP spread spectrum communication system, for example, the mobile station must find the boundaries of the downlink elementary signal, the symbol boundaries, and the frame boundaries of the clock reference clock signal generator.

длиной N_p элементарных сигналов. Код синхронизации может также определяться, как расширяющий код для символов, маскированных длинным кодом. Этот код синхронизации передается с известной модуляцией и без какого-либо скремблирования длинным кодом. В системе связи МДКР одного вида каждая базовая станция использует отличный, известный код синхронизации, взятый из набора доступных кодов синхронизации. В системе связи МДКР другого вида все базовые станции используют один и тот же код синхронизации, с различиями между базовыми станциями, которые идентифицируются по использованию для передачи различных сдвигов фазы кода синхронизации.The most common solution implemented to solve the synchronization problem, as shown in figure 1, implies the presence of a base station that periodically transmits (with a repeating period T _p ), and a mobile station that detects and processes the recognized synchronization code

length N _{p of} elementary signals. A synchronization code can also be defined as an extension code for characters masked by a long code. This synchronization code is transmitted with known modulation and without any scrambling with a long code. In one type of CDMA communication system, each base station uses a different, well-known synchronization code, taken from the set of available synchronization codes. In a different type of CDMA communication system, all base stations use the same synchronization code, with differences between the base stations, which are identified by the use of the synchronization code for different phase shifts.

At a spread spectrum receiver of a mobile station, the received signals are demodulated and used in a filter consistent with the synchronization code (s). It is understood that other detection schemes, such as moving correlation, may be used to process the synchronization code. The output of the matched filter has a short burst (peak) at times corresponding to the reception time of the periodically transmitted synchronization code. Due to the effects of multi-route transmission, individual peaks may be detected due to a single transmission of the synchronization code. When processing these received peaks in a known manner, the synchronization reference signal related to the transmitting base station can be determined with some error equal to the repetition period T _p . If the repetition period is equal to the frame length, then the synchronization reference signal can be used to synchronize the communication operations of the mobile station and the base station in accordance with the frame synchronization.

может быть выбрана любая длина N_p в элементарных сигналах, практически, длина N_p в элементарных сигналах ограничена сложностью согласованного фильтра, предусматриваемого в приемнике мобильной станции. В то же время, предпочтительно ограничить мгновенную пиковую мощность

передачи сигнала/канала кода синхронизации, чтобы не вызвать сильную мгновенную интерференцию с другими передаваемыми сигналами/каналами с расширенным спектром.While for the transmitted synchronization code

any length N _p in the chips can be selected; in practice, the length N _p in the chips is limited by the complexity of the matched filter provided in the receiver of the mobile station. At the same time, it is preferable to limit the instantaneous peak power.

transmitting the signal / channel of the synchronization code so as not to cause strong instantaneous interference with other transmitted signals / channels with a spread spectrum.

In order to provide a satisfactory average transmit power of a synchronization code with a certain given length N _{p of} chips, according to FIG. 2, it may be necessary in a CDMA communication system to use a synchronization code repetition period T _p shorter than the frame length T _f .

внутри одной длины кадра Т_f является, как известно специалистам в данной области техники, поддержка межчастотной синхронизации нисходящей линии связи в режиме уплотнения (сжатия). При обработке в режиме уплотнения, синхронизация нисходящей линии связи на данной несущей частоте осуществляется только на протяжении части кадра, а не на протяжении всего кадра. Тогда, при наличии только одного кода синхронизации

на кадр имеется возможность при отсутствии обработки в режиме уплотнения в течение значительного периода времени обнаружения полностью кода синхронизации. При передаче многих кодов синхронизации

за каждый кадр, для процесса обработки в режиме уплотнения предоставляется много возможностей для осуществления обнаружения за один кадр, и можно будет обнаружить, по меньшей мере, одну передачу кода синхронизации.Another reason for transmitting many synchronization codes

within one frame length T _f is, as is well known to specialists in this field of technology, support for inter-frequency synchronization of the downlink in the compression (compression) mode. When processing in compression mode, downlink synchronization at a given carrier frequency is carried out only for part of the frame, and not throughout the frame. Then, if there is only one synchronization code

per frame there is a possibility in the absence of processing in compression mode for a significant period of time to detect the full synchronization code. When transmitting many synchronization codes

for each frame, for the processing in compression mode, there are many possibilities for performing detection in one frame, and at least one transmission of the synchronization code can be detected.

внутри одного кадра длиной T_f. Принятые сигналы снова демодулируются и передаются к фильтру (или коррелятору), согласованному с известным кодом синхронизации. Выходной сигнал согласованного фильтра создает пики в моменты времени, соответствующие моментам приема периодически передаваемого кода синхронизации. При обработке этих пиков может быть найден известным образом опорный сигнал синхронизации для передающей базовой станции, связанный с периодом повторения Т_р кода синхронизации. Однако опорный сигнал синхронизации является неоднозначным в отношении синхронизации кадра и, таким образом, не предоставляет информацию, достаточную для обеспечения возможности синхронизации базовой/мобильной станции с опорным сигналом синхронизации. Под неоднозначностью понимается то, что обнаружение пиков кода синхронизации является недостаточным для идентификации границы кадра (то есть, его синхронизации).However, there is a definite drawback in receiving and synchronizing when transmitting multiple synchronization codes

inside one frame of length T _f . The received signals are again demodulated and transmitted to the filter (or correlator), consistent with the known synchronization code. The output of the matched filter generates peaks at times corresponding to the times at which a periodically transmitted synchronization code is received. By processing these peaks, the synchronization reference signal for the transmitting base station associated with the repetition period T _p of the synchronization code can be found in a known manner. However, the synchronization reference signal is ambiguous with respect to frame synchronization and thus does not provide information sufficient to enable the base / mobile station to synchronize with the synchronization reference signal. By ambiguity, it is understood that the detection of peaks in the synchronization code is insufficient to identify the frame boundary (i.e., its synchronization).

группы длинных кодов передается синхронно с кодами синхронизации

(и еще более предпочтительно, ортогонально к кодам синхронизации

). Код группы длинных кодов передается с известной модуляцией и без скремблирования длинным кодом. Каждый код

группы длинных кодов определяет конкретное подмножество общего набора длинных кодов, к которому принадлежит используемый для передачи определенный длинный код сотовой ячейки. Например, может иметься сто двадцать восемь общих длинных кодов, сгруппированных в четыре подмножества, каждое из которых состоит из тридцати двух кодов. Идентифицируя передаваемый код

группы длинных кодов, приемник, в данном примере, может сузить поиск извлечения своего длинного кода всего до тридцати двух длинных кодов, содержащихся в подмножестве, идентифицированном принятым кодом

группы длинных кодов.The cell search process may further involve obtaining a specific long cell code used on the downlink to scramble the data transmission of the dedicated and common downlink channels. Specialized channels include both traffic channels and control channels, common channels also include traffic channels and control channels (which may include a transmission control channel or a PMC). Preferably, according to figure 3, the code

groups of long codes transmitted synchronously with synchronization codes

(and even more preferably, orthogonal to the synchronization codes

) The code of a group of long codes is transmitted with known modulation and without scrambling with a long code. Every code

a group of long codes defines a specific subset of the general set of long codes to which the specific long code of the cell used to transmit belongs. For example, there may be one hundred twenty-eight common long codes, grouped into four subsets, each of which consists of thirty-two codes. Identifying the transmitted code

groups of long codes, the receiver, in this example, can narrow down the search for retrieving its long code to just thirty-two long codes contained in the subset identified by the received code

groups of long codes.

и кодов

группы длинных кодов. Мобильная станция сначала идентифицирует синхронизацию кода синхронизации, применяя к принятому сигналу

-согласованный фильтр и идентифицируя пики. По этим пикам может быть найден опорный сигнал синхронизации относительно интервалов времени. Определенные, хотя неоднозначно по отношению к синхронизации кадра, расположения интервалов времени идентифицируют синхронизацию для одновременной передачи кода

группы длинных кодов. Далее выполняется корреляция с известными расположениями интервалов времени для идентификации кода

группы длинных кодов. По этой идентификации уменьшается число возможных определенных длинных кодов сотовой ячейки, используемых для осуществления передачи. Наконец, в каждом из известных интервалов времени по каждому из уменьшенного количества длинных кодов (то есть, рассматриваются те длинные коды, которые содержатся в идентифицированном подмножестве

) выполняется корреляция, чтобы определить, определенный длинный код какой из сотовых ячеек используется для осуществления передачи, и обеспечить опорный сигнал сдвига фазы. Как только найден сдвиг фазы, синхронизация кадра является идентифицированной.Frame synchronization information can be determined by combined processing of received synchronization codes

and codes

groups of long codes. The mobile station first identifies the synchronization of the synchronization code by applying to the received signal

-consistent filter and identifying peaks. From these peaks, a synchronization reference signal with respect to time intervals can be found. Certain, albeit ambiguous with respect to frame synchronization, time slot locations identify synchronization for simultaneous code transmission

groups of long codes. Next, correlation is performed with known locations of time intervals for identifying the code.

groups of long codes. By this identification, the number of possible specific long cell codes used for transmitting is reduced. Finally, in each of the known time intervals for each of the reduced number of long codes (that is, those long codes that are contained in an identified subset are considered

) a correlation is performed to determine the specific long code of which of the cells is used to transmit and provide a phase shift reference signal. Once a phase shift is found, frame synchronization is identified.

внутри одного кадра длиной Т_f в патенте US №5991330 под названием "Синхронизация мобильной станции внутри системы передачи данных с расширенным спектром" рассмотрен альтернативный вариант способа определения синхронизации кадра путем наличия в каждом из интервалов времени не только кода синхронизации

, описанного выше согласно фигуре 2, но также и кода кадровой синхронизации

, передаваемого с известной модуляцией и без скремблирования длинным кодом как иллюстрируется фигурой 4. В каждом интервале времени и по всем повторяющимся кадрам используется один и тот же код синхронизации. Коды кадровой синхронизации, однако, являются уникальными для каждого интервала времени в кадре и повторяются в каждом кадре.Regarding the transmission of many synchronization codes

inside one frame of length T _f in US patent No. 5991330 entitled "Synchronization of a mobile station within a spread spectrum data transmission system" an alternative method for determining frame synchronization by the presence of not only a synchronization code in each time interval is considered

described above according to figure 2, but also the frame synchronization code

transmitted with known modulation and without scrambling with a long code as illustrated in Figure 4. The same synchronization code is used in each time interval and for all repeated frames. Frame synchronization codes, however, are unique for each time interval in a frame and are repeated in each frame.

-согласованный фильтр к принятому сигналу и идентифицируя пики. По этим пикам может быть найден опорный сигнал синхронизации относительно интервалов времени. При том, что опорный сигнал синхронизации является неоднозначным относительно синхронизации кадра, информация о расположении интервала времени косвенно указывает на расположение кода кадровой синхронизации

внутри каждого расположенного интервала времени. Далее мобильная станция дополнительно коррелирует набор известных кодов кадровой синхронизации

с принятым сигналом в местах расположения кодов кадровой синхронизации. При условии, что позиция каждого кода кадровой синхронизации

относительно границы кадра известна, как только в определенном расположении найдено соответствие корреляции, граница кадра, относящегося к нему (и следовательно, синхронизация кадра) становится также известной.To obtain frame synchronization information, the mobile station first identifies the synchronization of the synchronization code using

- A consistent filter to the received signal and identifying peaks. From these peaks, a synchronization reference signal with respect to time intervals can be found. While the reference synchronization signal is ambiguous with respect to frame synchronization, information about the location of the time interval indirectly indicates the location of the frame synchronization code

inside each located time interval. Further, the mobile station further correlates a set of known frame synchronization codes

with the received signal at the locations of the frame synchronization codes. Provided that the position of each frame synchronization code

relative to the frame boundary is known, as soon as a correlation is found in a specific location, the border of the frame related to it (and therefore frame synchronization) becomes also known.

Although previous methods for acquiring synchronization information may provide satisfactory results, their performance under poor radio conditions is poor. Inevitably, in all the approaches described above, corresponding to the prior art, insufficiently good radio link conditions and an increased level of interference can cause the mobile station to make the wrong decisions regarding the long code or frame synchronization, or both. Therefore, additional correlations are required that take up valuable resources of the processing process, which is difficult to implement, while the process of searching for a cell is slowed down. Essentially, one could obtain a greater signal power by receiving a signal over a larger number of frame periods. However, this approach may require more time than is considered acceptable for communication switching (forwarding) situations. Therefore, an effective method of obtaining and indicating frame synchronization and indicating a long code during a cell search process in a radio communication with deteriorated conditions is required. As described in detail below, the present invention provides such a method.

SUMMARY OF THE INVENTION

Essentially, the present invention provides a method for more efficiently extracting a long code and frame synchronization during cell search by using a special coding scheme. A code set having certain properties is determined, consisting of Q-th code words of length M, including characters from a set of Q short codes. The first property that is required to be ensured is that no cyclic shift of the codeword results in a correct codeword. Other properties that must be ensured are a one-to-one correspondence between the long code message and the correct code word, and the decoder should, in the presence of interference and noise, provide an arbitrary shift with a certain degree of accuracy and an acceptable degree of complexity (thus implicitly determining frame synchronization) and a transmitted codeword (i.e., a long code indication message relating thereto).

An important technical advantage of the present invention is that it provides a simple solution for more efficient long code recognition and frame synchronization during cell search, which provides coding gain that can be used to reduce search time and / or to provide the required information bit power ratio to noise (Mb / W).

Another important technical advantage of the present invention is that it provides the opportunity to find a compromise between complexity and efficiency by changing the complexity of the code, while keeping the number of possible messages fixed.

Another important technical advantage of the present invention is that it provides more code words than standard schemes that reduce and / or limit the number of required messages of the base station.

BRIEF DESCRIPTION OF THE DRAWINGS

For a more complete understanding of the method and device proposed by the present invention, the following is a detailed description of embodiments of the present invention according to the accompanying drawings.

1 is a diagram illustrating a signal transmission format in a synchronization channel in a code division multiple access communication system with a forward sequence code according to the prior art.

2 is a diagram illustrating an alternate embodiment of a signal transmission format in a synchronization channel in a code division multiple access communication system with a forward sequence code according to the prior art.

3 is a diagram illustrating another alternative embodiment of a synchronization channel and a long code group signal transmission format in a code division multiple access communication system with a forward sequence code according to the prior art.

4 is a diagram illustrating another alternative embodiment of a transmission format of a synchronization code and a frame synchronization code in a code division multiple access communication system with a forward sequence code according to the prior art.

5 is a diagram illustrating a possible embodiment of a transmitter and a receiver that may be used to provide the present invention.

FIG. 6 is a flowchart illustrating a generalized decoding algorithm that can be used in a receiver decoder to decode a codeword described above in FIG. 5 in accordance with a preferred embodiment of the present invention.

FIG. 7 is a flowchart illustrating, according to a second embodiment of the present invention, a generalized maximum likelihood decoder algorithm that can be used in a receiver decoder to decode a codeword described above in FIG. 5 in the presence of random symbol / bit interference / noise.

Figure 8 is a diagram that illustrates part of a possible variant of the lattice for trellis encoder, where m = 2, while this diagram is provided for explanatory purposes to illustrate the present invention.

Figure 9 is a schematic diagram illustrating a possible embodiment of an annular lattice with M equal to 8, wherein this schematic diagram is provided for explanatory purposes to explain the present invention.

Figure 10 is a block diagram of a possible algorithm that can be used by an encoder to generate all words that satisfy property 2 of the present invention.

Figure 11 illustrates a synchronization code that can be obtained by implementing the coding algorithm described in accordance with Figure 10.

FIG. 12 is a flowchart illustrating a method proposed by the present invention for decoding a possible embodiment of the synchronization code described above with respect to FIG. 11.

Figure 13 shows the source code of the Matlab program for the synchronization code search algorithm described according to Figure 12.

Figure 14 shows the cardinal number of some synchronization codes for small M.

Figures 15A and 15B are functional diagrams of possible trellis encoders that can be used to implement the present invention.

16 is a diagram illustrating a method for searching for a cell to be used by a mobile station, as described in ARIB's previously developed CDMA broadband communications offerings.

17 is a table illustrating some characteristics of a primary synchronization code and a secondary synchronization code.

Figure 18 is a table that provides information that, according to the present invention, may be provided with a primary synchronization code or a secondary synchronization code for searching a cell.

FIG. 19 is a possible embodiment of a method for providing the primary synchronization code and the secondary synchronization code shown in FIG. 18 for searching for a cell according to the present invention.

FIG. 20 is a second possible embodiment of a method of providing the primary synchronization code and secondary synchronization code shown in FIG. 18 for cell search according to the present invention.

Figure 21 is a table describing two algorithms (methods) for searching a cell, which can be used to implement the present invention, while this table also provides a comparison of two possible variants of a method for searching a cell proposed by the present invention, with a method for searching a cell according to the proposals ARIB for CDMA Broadband Communication Systems.

DETAILED DESCRIPTION OF THE DRAWINGS

The following describes a preferred embodiment of the present invention and its advantages, illustrated by drawings 1-21, which use end-to-end numbering.

Essentially, according to the present invention, a method for more efficiently extracting a long code and frame synchronization during cell search using a special coding scheme is proposed. A code set possessing certain properties is determined, consisting of Qth code words of length M, including characters from a set of Q short codes. The first property that is required to be ensured is that no cyclic shift of the codeword results in a correct codeword. Other properties that should be ensured are the presence of a one-to-one correspondence between the long code message and the correct code word, and the decoder should, in the presence of interference and noise, provide an arbitrary shift with a certain degree of accuracy and an acceptable degree of complexity (thus implicitly determining frame synchronization), and the transmitted codeword (i.e., a long code indication message relating thereto).

More specifically, to explain the present embodiment, suppose that the transmitter transmits M characters selected from the Qth alphabet (for example, an alphabet containing Q orthogonal short codes of length N). These transmitted symbols constitute a transmitted codeword, and a set of length M Qth sequences (codewords) can be defined as a code. In addition, the same codeword is transmitted several times.

The receiver (transmitted codewords) has information about the start and end time of the character, but does not have information about the start and end time of the codeword. Also, the transmitted signal gradually fades, is affected by interference, and / or noise. The receiver’s task is also (1) to extract the transmitted codeword (and the corresponding message) if there is no information regarding the start / end time of this codeword, and (2) to extract the start / end time of the codewords. 5 is a diagram illustrating the transmitter and receiver operations described above.

According to figure 5, the transmitted characters are denoted by a, b, c, ..., etc. It should be noted that in this example the characters a, b, c, d due to the frequency of the transmitted signal are respectively equal to the characters f, g, h, i. Also, it should be noted that any set of M consecutive characters contains all the information necessary for the receiver to decode the received signal, assuming that the receiver has information about the synchronization of the code frame. In this example, M is 5. If there is no information about the synchronization of the code frame, then the decoding process is nontrivial. However, in this example, for simplicity, it is assumed that there is information about the synchronization of the code frame, along with the use of code that has certain known properties. At the receiver (RX), any of a set of consecutive symbols at M = 5 offsets may contain the information necessary to decode the received signal.

In the following description, for simplicity, it can be assumed that the time interval between characters is zero. Also, to ensure that damaged characters corresponding to the codeword are retrieved with an acceptable degree of reliability, it can be assumed that a standard decoding method is used.

A channel (from the receiver's point of view) can be described as the introduction of arbitrary symbol errors arising from interference and noise, which can shift codewords by a random number of (full) Qth symbols. The transmitter retransmits the same message several times. Therefore, any M received consecutive characters (regardless of their position) can be a codeword corresponding to an unknown cyclic shift. Also, a code set consisting of Qth code words of length M (with characters from a set of Q short codes) is determined by the following properties.

Property 1: There is a one-to-one correspondence between the long code message and the correct code word (there are L code words or messages).

Property 2: no cyclic shift (of the Qth characters) of the code word results in the correct code word (if the number of shifts is not zero or a multiple of M, which is a trivial solution).

Property 3: the decoder should, with some degree of accuracy and, preferably, at an acceptable degree of complexity, provide a random shift (thus implicitly determining frame synchronization), and the transmitted codeword (that is, the corresponding long code information or IDK message) in the presence of interference and noise. Especially, as described in detail below, a preferred embodiment of the present invention uses codes that primarily provide property 2. Also, as described below, it follows that these codes also satisfy Properties 1, 3.

First, to further facilitate understanding of the present invention, consider a (simplified) channel in which bit / symbol errors do not occur, and only an unknown number of cyclic symbol shifts of an unknown, repeatedly transmitted codeword occur. The receiver must determine both the actual offset and the transmitted codeword.

6 is a flowchart illustrating, according to a preferred embodiment of the present invention, a generalized decoding algorithm that can be used in a receiver decoder to decode a codeword described above (although this algorithm is not the most efficient decoding method). At operation 101, the decoder collects M consecutive characters (received word). Next, in operation 102, the decoder determines whether the received word is the correct code word. If not, then the decoder performs operation 103. Otherwise, the decoder performs operation 104.

Also, if the received word is not a valid code word, then at operation 103, the decoder cyclically shifts the received word by one step (symbol), and then returns to operation 102. Or, at operation 104, the decoder displays the number of shifts made during operation 103 performed to obtain the correct codeword, and thus receives a message corresponding to that codeword. The number of shift output in operation 104 gives the synchronization of the codeword frame.

7 is a block diagram illustrating, according to a second embodiment of the present invention, a generalized maximum likelihood decoder algorithm that can be used in a receiver decoder to decode the codeword described above in the presence of symbol / bit interference / noise (however, this approach also does not is the most effective). The generalized maximum likelihood decoder algorithm provides an estimate or probability regarding how likely the received bits are and are actually a codeword. An estimate or probability is represented as a probability value (likelihood). In this possible embodiment, the present invention provides for assembling k · M symbols before actually decoding the received word, which provides a more correct estimate of the decoded word than the first method described above, since this produces a plurality of (k) instances of all code symbols.

When a possible variant of the maximum likelihood algorithm provided for by this embodiment is used in the decoder, in step 201, the decoder collects k · M consecutive characters (received word) and combines the values of the probabilities of the characters. At operation 202, for each of the L codewords and each of the M cyclic symbol shifts, the decoder calculates the correlations between the received word and the corresponding combinations of L codewords from their M character shifts. The decoder stores both the codeword and the number of necessary shifts that lead to the best correlation. At operation 203, the decoder outputs the stored codeword (or corresponding message) and the number of offsets that led to the best correlation.

According to a third embodiment of the present invention, a more efficient decoding algorithm will be described together with an example illustrating the existence of codes that may satisfy properties 1-3. Essentially, according to this possible embodiment, the presented decoding algorithm combines the so-called trellis code of the tail bits and the synchronization code, such that all properties 1-3 are satisfied. Also, the code is constructed by sequentially connecting (concatenating) the internal trellis code with an external code having synchronization properties, so that the constructed general code satisfies property 2. It follows that properties 1 and 3 will also be satisfied.

Specifically, according to this possible embodiment of the present invention, an input binary tail trellis internal binary encoder that produces Qth characters is first considered (by way of example only). These symbols can be a complex scalar or complex vector signal. It is assumed that an input binary frame of length M bits is provided. It is assumed that the input, initial state in which the encoder must be in order to complete actions in the same state can be calculated as follows. For polynomial encoders of order M, the initial state can be set equal to the last M bits in the input frame. Therefore, the encoder and decoder start and end the action in the same state. However, this state is unknown to the decoder. Also, the correct code words are words that can be obtained by moving around the grid, starting from a certain state and ending in the same state corresponding to the initial state.

Figure 8 is a diagram showing a possible embodiment of a grating section for m = 2 trellis encoder, which is provided for explanatory purposes. The four rectangles placed vertically on the right side of Figure 8 represent the four possible states of the shift register for the trellis encoder at m = 2, the contents of which are indicated inside these rectangles. The complete lattice contains M series-connected (concatenated) sections identical to the lattice sections shown in figure 8.

For the trellis tail bit encoder, the lattice goes around and the last state column becomes the same as the first. Arrows with labels (for example, I / code 1) indicate that, given the current state of the encoder (the state from which the arrow exits) and the input signal (I), the current output symbol is code (1), and the next state will be the arrow points to. It should be noted that the arrows shown in FIG. 8 have particular marks, but this is not limited to the present invention. The code labels shown in the figure are provided for explanatory purposes only and are not intended to determine the specific matching functions used.

As explained by the lattice section shown in FIG. 8, all phases of the lattice in the lattice code are identical, and the same code is repeated many times. Therefore, the codeword path can be considered as the path in the ring lattice shown in Figure 9. Also, Figure 9 is a schematic diagram of a possible variant of the ring lattice with M equal to 8. Each rectangle shown is a state column (for example, like one of the right / of the left columns shown in figure 8), and each arrow shown indicates a set of possible changes (transitions) of states and the corresponding input-output ratio. As mentioned previously, according to the present invention, all shown lattice phases are identical. Therefore, any cyclic shift of the output symbol sequence is also a valid output sequence. Also, the start state and the end state in the path of the annular lattice are the same, but the actual position in the lattice where the start / end state is is unknown.

The decoder used collects M consecutively received symbols and makes an assumption about the position of the start / end state in the lattice. All cyclic shifts of admissible trajectories are also admissible trajectories. Therefore, the correct path (but not the start / end position) can be decoded (assuming that the noise level is not too high). It should be noted that although this tail-bit encoder algorithm does not satisfy property 2 (above), all symbol-wise cyclic shifts of the codeword are valid codewords. However, the use of such a lattice structure easily allows the use of programmable decoding methods and structured lattice diagrams for more efficient decoding. A brief overview of known methods that can be used to decode trellis codes of tail bits is given in the work of R. Koch and S.-E. Sandberg, "Effective Viterbi Adaptive Ring Algorithm for Decoding Generalized Convolution Codes of Tail Bits," IEEE Proceedings on Transport Technologies, vol. 43, number 1, 1994, and in US patent No. 5355376. Thus, assuming that the correct path has been decoded (the most probable case), a variant of the ring shifts of the input M bit frame can be obtained.

In order to limit the words of the internal code described above so that property 2 is satisfied, an external synchronization code of length M bits is introduced, which makes up the said M bit frame. As described below, the external synchronization code satisfies property 2. Therefore, if we consider both the internal and external codes as a single code, then this resulting single code satisfies property 2.

After the internal code has been decoded, a shifted version of the external code can be obtained. However, only one exact shift of this decoded word produces the correct external codeword. Therefore, the word obtained after decoding the inner code is shifted until the correct code word is received. The number of required shifts determines the frame synchronization and the message corresponding to the IDK. If the correct codeword does not appear after performing M offsets, it can be concluded that an internal decoding error has occurred, thus the present invention provides a form for detecting errors.

The following illustrates the fact that such synchronization codes (which satisfy property 2) do exist, and codewords in the external code are listed for small values of M. Also, a trellis code is defined for a number of different embodiments.

A series of possible synchronization codes (and their cardinal number) such that each code can provide property 2 according to the present invention is described herein. In this example, for explanatory purposes, M is set equal to 5, but further considerations are also valid for any value of M. According to property 2, the restriction imposed on codewords is that any (non-trivial) cyclic shift should result in an excellent not -the codeword. Also, the "period" of a word is defined as the number of cyclic shifts needed to return to that word. In this embodiment, the period is assumed to be less than or equal to M. The "P-cycle" is defined as the set of "p" words with the period "p", which is obtained by shifting the "p" of the words of the period. The restriction placed on each codeword is that it has a period M and that M-1 shifts do not result in codewords. Under the conditions defined above, the encoder can use the following algorithm, illustrated in figure 10, to generate all words that satisfy property 2.

According to figure 10 at operation 301, for all 2 ^M words of length M, the encoder calculates the period of these words. Then, in operation 302, the encoder excludes from consideration all words with periods less than M. In operation 303, the encoder excludes all words in M-cycles except one that can represent a cycle (for example, the smallest if the word is treated as a binary number). At operation 304, the encoder assumes that the remaining words satisfy property 2 and constitute the code of interest. An illustration of the above algorithm is shown in Figure 11. As shown, M is 5. The arrows pointing to the right (->) indicate that there has been a cyclic shift (for example, the right one). All 2 ⁵ = 32 words are taken into account, and six words remain in the resulting synchronization code (rightmost column). Therefore, in the example considered, the synchronization code of interest contains six codewords 1, 3, 5, 7, 11, 15 (decimal), and therefore L = 6.

FIG. 12 is a flowchart illustrating a method of decoding a possible embodiment of the synchronization code described above according to FIG. 11 in accordance with the present invention. At operation 401, the decoder collects M consecutive bits (obtained after internal decoding). At operation 402, the decoder shifts the received frame until this frame becomes as small as possible (for example, takes the form of a binary number), a maximum of M times. At operation 403, the decoder determines whether the resulting word is a codeword. If so, then at step 404, the decoder outputs a message corresponding to the codeword along with the number of offsets that were necessary to obtain the codeword. Otherwise, if the resulting word is not a code word, then an assumption may be made that an error occurred during internal decoding. In this case, at operation 405, the decoder may output an internal decoding error message. Figure 13 shows a listing of the above algorithm for finding the synchronization code, and figure 14 shows the cardinal number (that is, determining the number, but not the order) of some synchronization codes for small M.

In a CDMA-PP system, M code symbols may contain a number Q of so-called short codes of length N. These short codes are often orthogonal to each other, or otherwise have good cross-correlation characteristics. Consider a low-speed, time-independent trellis code in which the symbols on the branches of the lattice are vectors that are selected from a set of Q vectors defined above (or simply symbols, as called here). For example, a similar set of vectors is described in US patent No. 5193094.

Figures 15A and 15B are functional diagrams of possible trellis encoders 10 and 20, respectively, which can be used to implement the present invention. Essentially, such a trellis encoder is constructed in the form of a shift register (12, 22) of length m, with an input signal, I, and a converter (14, 24), which performs state conversion of the corresponding shift register (12, 22) and the current input signal, I, into the output vector (for example, c1, c2, ..., cN). In trellis encoders 10, 20, as shown in the figure, the length of the shift register (m) is 3. Therefore, each of the shift registers 12, 22 can take 8 different states. The set of required vectors / symbols (e.g., c1, c2, ..., cN) makes up the set of orthogonal vectors for the orthogonal trellis encoder 10 (Figure 15A), and the set of orthogonal or diametrically opposite vectors for the superorthogonal trellis encoder 20 (Figure 15B).

Essentially, an orthogonal lattice code is obtained if the transformation of the state of the register and the input signal, I, results in a vector, and if the set of vectors thus obtained forms a set of orthogonal vectors. A superorthogonal code is generated if the first m-1 register states define an orthogonal vector, and can be taken as an output vector if the sum modulo 2 of the input bit and the state content of the mth register is not equal to 1. In this case, the output vector is bitwise inverted by the inverter 26. With a standard conversion, such as 0/1 -> + 1 / -1, you can see that the output for some state is diametrically opposite vectors depending on the input 0 and 1, respectively. Essentially, such codes are suitable for use in CDMA-PP, such codes can be used as symbols due to their inherent expansion effect (very low code speed), good correlation characteristics, and their inherent error correction capabilities due to lattice structure.

In addition to the new encoding (decoding) method described above according to the present invention, a new method is also proposed for using a frame synchronization indicator (USC) to search for a cell using the coding scheme proposed by the present invention, for example, a cell search scheme according to ARIB proposals for a broadband communication system CDMA Essentially, the downlink channels described in current ARIB CDMA broadband offerings facilitate the three-step information extraction process used by the mobile station. However, since the mentioned channels related to information extraction do not contain information on frame synchronization, the final stage of the proposed ARIB procedure is rather complicated and / or time-consuming. As described below, the present invention provides at least two methods that can be used to provide USK, for example, within a circuit structure proposed by ARIB over a CDMA CDMA communication system.

Specifically, FIG. 16 is a diagram illustrating a method for searching for a cell to be used by a mobile station according to ARIB's proposals for CDMA broadband communication systems. In each time interval, the primary synchronization code (PCN) and the secondary synchronization code (VKS) are transmitted in parallel, both with known modulation, but without scrambling with a long code. The length of the PKS / VKS is one character of a physical channel transmitting 16 kilosymbol / s, or 256 elementary signals. The system has N _VCS VCS permissible, which give log ₂ (N _FSI) bits of information to be used for the IIR. The characteristics of the PKS and VKS are summarized in the table shown in figure 17. As explained in figure 17, no USK is provided for the mobile station, which can lead to a much longer search for a cell than is necessary.

Figure 18 is a table that provides information that can be provided by the PKS / VKS for cell search according to the present invention. Although this information can be provided in several ways, two possible embodiments are described below that can be used for the cell search schemes provided herein.

Specifically, according to one embodiment of the present invention (as explained in figure 19), as in the proposed ARIB scheme, VCS is the same in each slot in the frame, and the system has N _VKS allowable VCS, which give log ₂ (N _FSI) information bits which should be used for IDK. VKS throughout the frame is additionally modulated by one of the N _{MODs of} possible variants of a valid (for example, binary) sequence of length 16. This method offers IDK and other log ₂ (N _MOD ) bits of information for using IDK. The resulting modulating sequences of length 16 have good autocorrelation characteristics.

If the N _MOD value is greater than 1, then the following properties must also be ensured: (1) good cross-correlation and (2) the cyclic shift of any correct modulating sequence cannot lead to another correct modulating sequence (and any cyclic shift). If the modulating sequences obtained in this way satisfy the above properties, then the USK becomes known when any valid modulating sequence is detected in the receiver of the mobile station. Coherent detection of the received signal is facilitated by the use of the PCS as reference symbols to obtain a reference signal of the channel phase. Also, typical of USK. Therefore, all log ₂ (N _VKS ) + log ₂ (N _MOD ) bits of information can be used for IDK.

According to a second embodiment of the present invention (as illustrated in FIG. 20), there is a sequence of 16 video conferencing units that is repeated in each frame. In general, there are N _VKS _ _last similar VKS sequences that can be used in a system producing log ₂ (N _VKS _ _last ) bits of information that can be used for the IDK. In this case, it is preferable that each VKS sequence is unique and the individual VKS have good autocorrelation and cross-correlation characteristics. However, it can be assumed that in practice the equality N _VKS _ _last = 1 will be fulfilled.

Finding the right VKS sequence essentially yields the USK, and the VKS sequence can also be modulated, according to the method described immediately above for the first embodiment, which produces log ₂ (N _MOD ) bits of information for using the IDK. In this case, the IDK can use 65,536 different values (more than enough), which provide greater detection efficiency of the IDK.

Figure 21 is a table that explains two cell search algorithms (methods) that can be used to implement the present invention. Also, the table shown in FIG. 21 provides a comparison of two possible variations of the cell search method according to the present invention with the current cell search method according to ARIB proposals for a CDMA broadband communication system. The table rows (steps) in FIG. 21 illustrate the included cell search steps. For example, in the first stage (stage 1), a matched filter (SF) is used to ensure synchronization of time intervals (SII). In the second stage, when correlating (CORR) with the VKS, since the PCB provides a phase indicator, correlations can accumulate coherently. On the other hand, correlations can only be performed once per time interval, because there is only one VKS per time interval. When correlating with a long code (DC) in the third stage, correlations must accumulate incoherently. However, this correlation can be performed on consecutive characters, since a long code is applied to each character in the frame. In this case, the correlation is performed by sequentially connecting a long code and a known short PCC code, which is always transmitted in a downlink. If the long code can be accurately indicated by the IDK, then only one correlation step is necessary according to the two possible embodiments described above. However, with the currently proposed ARIB cell search scheme, in addition to the steps described above, an additional search is required to obtain frame synchronization (SC).

To illustrate, by way of example, receiver operations necessary for implementing the methods illustrated in Figure 21, the following samples can be made: N _VKS = 256 long codes, grouped as 16 × 16, N _MOD = 1, N _VKS = _last = 1, it is assumed (for simplicity) that the coherent accumulation of 16 correlations (256 chips each) is sufficient for adequate detection. To ensure the currently proposed ARIB cell search scheme, the following correlation matrix is formed:

обозначают 16 различных ВКС,

обозначают 16 последовательно принятых ВКС, и скалярное произведение обозначает, что корреляция осуществлена. При наличии 16 корреляторов в приемнике мобильной станции, 16 корреляторов должны использоваться для 16 интервалов времени, чтобы сформировать 256 корреляций Z₁. Элементы Z₁ могут также быть умножены на соединения соответствующих ПКС корреляций для устранения сдвига фазы, получившегося в результате ошибок синхронизации частоты и канала радиосвязи. Можно допустить, как то, что это умножение уже выполнено в матрице (1) выше, так и выполнение умножения во время описываемых ниже этапов. Затем строки Z₁ суммируются. Одна из этих сумм будет иметь большее значение, чем остальные, которое и определяет ВКС.Where

designate 16 different videoconferencing,

denote 16 consecutively adopted VKS, and the scalar product indicates that the correlation is implemented. If there are 16 correlators in the receiver of the mobile station, 16 correlators must be used for 16 time intervals to form 256 correlations Z ₁ . Elements Z ₁ can also be multiplied by the connections of the corresponding correlation corrections to eliminate the phase shift resulting from synchronization errors of the frequency and the radio channel. It can be assumed that this multiplication has already been performed in the matrix (1) above, and that the multiplication is performed during the steps described below. Then the lines Z ₁ are summed. One of these amounts will be more important than the rest, which determines the videoconferencing.

According to the first embodiment (method 1 described above), a matrix (1) is also formed. However, in order to implement method 1, matrix (1) is further multiplied by the following matrix:

where the columns contain all the cyclic shifts of the modulating sequence (here for simplicity it is assumed that these are real values). Multiplication Z ₁ M ₁ results in a 16 × 16 matrix, where one of the elements will have a larger value than the others. The row index of this element defines the IDK, and the column index determines the frame synchronization (USC).

According to a second embodiment of the present invention (method 2 described above), instead of matrix (1), the following matrix is formed:

являются ВКС из последовательности ВКС. Матрица (3) затем умножается на следующую матрицу:Where

are videoconferencing from the sequence of videoconferencing. Matrix (3) is then multiplied by the following matrix:

where the columns represent all 16 possible modulating sequences (again, real values are assumed for simplicity). Multiplication of the Z ₂ M ₂ matrices again results in a 16x16 matrix, where one of the elements should have a greater value than the others. The row index of this element defines USK, and the column index determines the IDK.

The operations for the above methods of the present invention can be extended to include more general cases. For example, if more modulating sequences are required, then the matrix M ₁ (M ₂ ) can be expanded with new columns containing all the shifts of all possible m-sequences. If more VKS sequences are required when performing the second method described above, then the matrix Z ₂ can be expanded by adding rows of shifted correlations with all possible VKS sequences. If there are more long codes per group, then the matrix, Z ₁ described above for the first method, can be expanded by adding more correlation lines. Essentially, with a limited set of correlators for use, correlations can be performed in successive frames, while their coherent accumulation continues. This observation in accordance with the present invention is valid for both of the cell search methods described above.

The following is a comparison of the two cell search methods of the present invention with a cell search scheme according to ARIB proposals for a CDMA broadband communication system. To make such a comparison, it is assumed that the system parameters shown in Table 1 below are applicable for each of the following cases.

Table 1 Parameter Value Chip rate 4.096 ms / s The speed of the physical channel symbol carried by the PMC 16 kilograms of characters / sec Frame length 10 ms Time intervals per frame 16 Symbols for the time interval 10 Elementary signals per symbol 256 The number of correlator blocks in the MS 16 The number of coherently accumulated correlations of 256 chips necessary for sufficient noise / attenuation suppression 16 The number of non-coherently accumulated correlations of 256 chips needed to sufficiently suppress noise / attenuation 32 The number of long codes in the system 256 Long Code Grouping 1 × 256, 4 × 32,
16 × 16, 32 × 4

The following tables 2-5 illustrate the advantages of the two cell search methods proposed by the present invention over the cell search method according to the ARIB proposals for CDMA broadband communication systems. For example, table 2 below shows the number of required correlations of 256 chips and the time required to obtain downlink synchronization for the three cell search schemes for the case where the long code grouping is not contained.

Table 3 below shows the same information for the case where there are four long code groups, each of which contains 32 codes.

Table 4 below shows the same information for the case where there are 16 groups of long codes, each of which contains 16 codes.

Table 5 below shows the same information for the case where there are 32 groups of long codes, each of which contains 4 codes.

The first stage (matched filtering or SF) is the same for all three methods. Therefore, this step is omitted from Tables 2-5 above for ease of presentation. For some correlations, the maximum and average values are given. The reason for this approach is that when a search is performed with insufficient information (blind search) for DC and SC, the correlation process can be completed before all possible combinations are found, if a good enough match is obtained. When performing blind searches (for example, among N different codes), on average N / 2 codes should be conducted. However, for the worst case scenario, you will have to check all N codes. Also, it may be decided to perform matrix multiplication, Z ₁ M ₁ , immediately, and their complexity is thus not taken into account in the tables above. In conclusion, as explained in Tables 2-5, two possible variants of the cell search method described above and performed according to the present invention facilitate, make it faster and less complicated, the cell search process in the mobile station, including the initial synchronization, and during situations reporting transmission characteristics. Also, as Tables 2-5 above show, both the delay and complexity of the cell search methods of the present invention are lower than for the cell search method according to the proposals of ARIB. In particular, the third stage (step 3) of the cell search procedure in the mobile station, performed according to the two methods provided by the present invention, is 16 times faster and less complicated than in the method proposed by ARIB.

Although the preferred embodiment of the method and device proposed by the present invention was illustrated and described in the foregoing detailed description, it is understood that the present invention is not limited to the considered embodiments, but can be modified in various ways by permutations, changes and substitutions, without departing from the gist invention, which is set forth and defined by the claims.

Claims (19)

1. The encoding method at the base station of at least one codeword for transmission, characterized in that they generate a set of identification code at the base station, transmit at least one codeword included in the identification code set from the base station, the identification code set comprising a plurality of codewords, each of which includes a plurality of characters taken from a set of short codes, wherein each codeword of a plurality of codewords is defined so that no character-wise cyclic s shift of each code word does not result in a correct code word. 2. The method according to claim 1, characterized in that the set of code words contains many Q-th code words, and the set of short codes contains a set of Q short codes. 3. The method according to claim 2, characterized in that the set of Q-th code words contains a lot of Q-th code symbols of length M. 4. The method according to claim 1, characterized in that the identification code is formed by sequentially connecting the internal and external codes. 5. The method according to claim 4, characterized in that the internal code contains a trellised code of the tail bits. 6. The method according to claim 4, characterized in that the external code contains a binary code. 7. The method according to claim 5, characterized in that the trellis code of the tail bits contains an orthogonal trellis code. 8. The method according to claim 5, characterized in that the trellis code of the tail bits contains a superorthogonal trellis code. 9. The method according to claim 2, characterized in that the short codes in the set Q of short codes are orthogonal short codes. 10. A method for a mobile station to decode an identification code transmitted from a base station in a code division multiple access (CDMA) cellular communication system, characterized in that a plurality of consecutive symbols containing an identifying code are received, it is determined whether the received plurality of consecutive symbols contains a correct code a word, and if the received set of consecutive characters does not contain the correct code word, then the received set of consecutive symbols is cyclically shifted ols value by a predetermined character and a return is made to the definition of the operation, if the received plurality of consecutive symbols comprises correct codeword, then output the number of cyclic shifts made to obtain the correct codeword and message corresponding to a correct codeword. 11. The method according to claim 10, characterized in that the number of cyclic shifts made to obtain the correct codeword determines the frame synchronization for the correct codeword. 12. The method according to p. 10, characterized in that the set of consecutive characters contains a predetermined number of consecutive characters. 13. The method according to claim 10, characterized in that the predetermined value contains one character. 14. A method for a mobile station to decode an identifier code transmitted from a base station in a Code Division Multiple Access (CDMA) cellular communication system, characterized in that k times M consecutive symbols are collected, while M consecutive symbols contain an identification code; the probability value for collected k times M consecutive characters, thus, get a set of M consecutive characters, calculate the correlation between each of L code words and each of M q -crystal shifts from a set M of total probability values and storing a code word and number of cyclical shifts made that led to the largest correlation value in calculating the correlation. 15. A method for a mobile station to decode an identifier code transmitted from a base station in a Code Division Multiple Access (CDMA) cellular communication system, characterized in that k times M consecutive symbols are collected, while M consecutive symbols contain an identification code; the probability value for collected k times M consecutive characters, thus, get a set of M consecutive characters, calculate the correlation between a set of M total probabilities by each of the M cyclic shifts L of the code words and store the code word and the number of cyclic shifts made, which led to the largest correlation value in calculating the correlation. 16. The method according to 14 or 15, characterized in that the number of cyclic shifts made determines the frame synchronization for the identifying code. 17. The method according to 14 or 15, characterized in that it further output a message corresponding to the stored codeword. 18. A method of encoding an identification code intended for transmission from a base station in a code division multiple access (CDMA) cellular communication system, characterized in that a period for each of 2 ^M words of length M intended to be encoded as an identification code is calculated, exclude each of 2 ^M words that has a period shorter than M, determine the given word for each of the M cycles of the remaining 2 ^M words, and save each given word. 19. The method of claim 18, wherein each of the stored predetermined words contains a plurality of codewords, and no cyclic shift of any of the plurality of codewords results in a correct codeword.

Images