RU2368104C2 - Distribution of resources in communication networks - Google Patents

Abstract

FIELD: physics, communication.

SUBSTANCE: invention is related to communication systems, in particular to distribution of communication resources in them. Method is suggested for adaptation of resource distribution parametres to achieve one or more quality indices with improved accuracy. New measurements of information are presented, based on so called mutual information (MI), preferably at block level. Measurements of information based on MI of the previous transmission, prediction of further transmission channel and one or more quality requirements are used to identify quantity and type of resources, for instance time resources, frequency and power, which must be used for further transmission. Distribution of resources may comprises, for instance distribution of power and/or adaptation of connection. Preferable version of realisation - mutual adaptation of connection and power distribution. Suggested method is useful in combination with automatic or hybrid automatic request of repeated transmission.

EFFECT: improved control of communication resources, improved reliability and efficiency of transfer.

40 cl, 10 dwg

Description

FIELD OF THE INVENTION

The present invention generally relates to communication networks and, in particular, to the allocation of resources in such networks.

State of the art

Today, a common problem with communication networks is the allocation of resources, such as transmission power and frequency, in an appropriate manner. Resources are limited, and as the number of connections and subscribers continuously increases, the network becomes more and more complex, which necessitates more complex solutions. As a rule, the goal in this case is the efficient use of resources and the reliability of transmission.

Automatic Repeat Request (AZP) and Hybrid Automatic Repeat Request (HARP) are widely used in data transmission to maintain transmission quality. The ALP retransmits the data blocks when a Reception Unconfirm (REC) message is received, indicating an incorrect reception. The receiver immediately discards the damaged blocks. Instead, the principle of HARQ is to buffer data blocks that were not received correctly and to combine buffered data with retransmissions. The combination software routine usually depends on what type of HAZP combining scheme is used, for example, Direct Combination (HAZP-PC) or Incremental Redundancy (HAZP-II) [1].

Existing solutions like the aforementioned AZP / HAZP mechanisms are associated with a number of problems. AZP / GAZP is trying to maintain the quality of the transmission, but cannot guarantee successful transmission. Even with the maximum number of retransmissions, a block may not be received correctly in a poor transmission medium. A larger limit on the maximum number of retransmissions will increase transmission reliability, but will require a larger buffer size and result in a longer transmission delay.

To achieve higher transmission efficiency, a number of studies of the planning of HARQ / Adaptive Modulation and Coding (AMC) based on channel prediction [2] were carried out. The main problem of the HAZP / AMK mechanism is the adaptation or counteraction to a change in the wireless channel, and the most common measure is to use the required average signal-to-noise ratio as an indicator. GAZP / AMK requires exhaustive modeling of all possible channel changes, which is an intractable task.

Accordingly, resource management during transmission in conventional communication systems is far from satisfactory, and there is a significant need for an improved resource allocation mechanism.

SUMMARY OF THE INVENTION

The overall objective of the present invention is to achieve improved resource management in wireless communication networks. A specific objective of the invention is to ensure transmission reliability and transmission efficiency. Another objective of the invention is to reduce the waste of resources in networks. Another objective of the invention is the provision of an improved mechanism for the allocation of resources for use with AZP / HAZP.

These objectives are achieved in accordance with the attached claims.

In short, the present invention provides a new method for adapting resource allocation parameters in order to achieve one or more quality goals with increased accuracy. In the allocation of resources, new quality indicators are presented based on the so-called mutual information (VI), preferably at the block level. VI-based quality indicators of a previous transmission, prediction of a subsequent transmission channel, and optionally one or more additional quality requirements are used to determine the amount and type of resources, for example, time, frequency, and power resources, which must be used for subsequent transmission. The resource allocation preferably comprises power distribution and / or link adaptation. The latter may, for example, include adaptive selection of the modulation mode, coding rate, and / or source data rate based on channel conditions. It also includes channel distribution among multiple users. In particular, the invention provides an opportunity for a useful implementation where connection adaptation and power distribution are performed simultaneously, based on the same measurement. The proposed method is very useful in conjunction with retransmissions of AZP / HAZP.

According to other aspects of the invention, there is provided a communication node and a communication system with means for allocating resources.

Brief Description of the Drawings

The invention, together with its further objects and advantages, will be understood with reference to the following description and the accompanying drawings, in which:

Figure 1 is a schematic view of a quality model for calculating a quality indicator according to an embodiment of the present invention;

2 is an illustration (parts) of a communication system in which the present invention can be used;

3 is a flowchart of a resource allocation method according to an embodiment of the present invention;

4A-4C are structural diagrams illustrating various mechanisms of a function defining a quality indicator in a communication device according to embodiments of the present invention;

FIG. 5 is a graph illustrating a mapping between an IPB and an SCE for a HARQ-PC according to an embodiment of the present invention;

6 is a graph illustrating display functions between IR and IPB for HARQ-II according to an embodiment of the present invention;

7 is a structural diagram of a system for allocating resources from HARQ-PC according to an embodiment of the present invention; and

FIG. 8 is a block diagram of a system for allocating resources with HARQ-II according to an embodiment of the present invention.

Detailed Description of the Invention

A list of abbreviations is provided at the end of this section.

The present invention includes the definition of new measurements of information and parameters, on the basis of which the distribution of resources (such as power distribution or adaptation of the connection) is performed on communication connections in communication networks. Before the invention and its embodiments are described in detail, some basic concepts and definitions will be explained in order to understand their principles.

New connection-system interface and information parameters

The present invention is based on the recognition of a new improved connection-system (CC) interface, which is also referred to as a quality model that provides practically optimal rules / requirements for resource allocation (with or without AZP / GASP), provided that they can be obtained Good estimates / predictions and channel measurements.

1 is a block diagram illustrating an embodiment of a similar quality model according to the invention. The quality model 100 describes a mapping relationship between measurements 11 of the connection information (OSB in FIG. 1) and the resulting quality indicators or ratings (BSEC and IR in FIG. 1). It contains a modulation model 12 and a coding model 13, respectively. As will be described below, an advantageous feature of the proposed quality model 100 is that it represents a linear interface between the modulation model 12 and the encoding model 13.

The demand for quality of service can be expressed through various quality indicators: BSBO (Error Block Frequency), throughput, delay, as well as one or more new indicators defined according to the invention. These indicators can be obtained statistically or based on measurements of the connection information, such as the SIR or the raw Frequency of Error Bits, and play an important role in the allocation of resources according to the invention. New quality indicators will be displayed in the section “Mutual information at the block level”, described below, but first, at the symbol level, the concept of mutual information on which the indicators are based will be explained.

Mutual information at a symbolic level

Seen by the decoder, information from the source is carried by the software outputs of the demodulator. The classical value of information from information theory is the so-called mutual information (VI) between the input and output of the channel, that is, between the output bit of the encoder and the input program bit of the decoder. The channel coding theorem states that an ideal codec (that is, an encoder / decoder system) is capable of reliably transmitting equal channel mutual information at a coding rate [3]. A measure of information based on channel capacity can be expressed as the value of modulated mutual information (VI) at the symbol level. For γ _j representing the signal-to-noise ratio (SIR) at time j, i.e.

,

,

SI is designated as I (γ _j ) and is defined as:

,

,

where the modulated symbol X belongs to a particular modulation group, and the accepted symbol Y = (Y _R + i * Y _I) ∈C , where C is a set of complex numbers [4]. In equation (2), P (X) is a priori probability X. P (Y | X, γ _j ) is a function of the probability distribution density of Y , due to the transmission symbol X and the parameterized state γ _{j of the} channel.

There are also approximations of symbolic information, which can sometimes be used instead of formula (2). For example, on the basis of the combined Chernoff limit for the encoded transmission, the expression for the Exponential Mapping of the Effective SIR (EOS of the SIR) for modulating an M-symbol has the following form:

,

,

where γ _m is the modulation correction factor for a given group.

Other approximate expressions of mutual information can, for example, be based on the cutoff frequency of binary phase shift keying (BPS):

or based on Shannon’s unified information theory, that is, real Gaussian inputs of additive white Gaussian noise (ABGS):

.

.

There may also be other approximate expressions, such as:

,

,

where {α, β} is the modulation compensation exponent for a given group. With good training, formula (6) gives a very good match.

Block Level Mutual Information

The characteristic of a particular codec can be expressed as mutual information per coding unit.

On the coding block (N, K), where K denotes the number of bits of information and N denotes the number of coded bits in one coding block, which corresponds to J modulated symbols, the channel capacity is the accumulation of SI in the block. Assuming that the received coded block is subject to a plurality of channel states {γ ₁ , γ ₂ , ..., γ _J }, the mutual information according to the present invention is further defined at various levels as the following quality indicators:

- Information received coded block (IPB):

For modulation in order M, I (γ _j ) ≤log ₂ M = N / J, and, therefore, IPB (γ _j ) ≤N.

where IPB denotes the information of the received interval , SI is the symbol information , that is, I and S are the number of characters in one interval.

IPB is the general information of the received coded (raw) bits before decoding.

- The frequency of information received coded block (CHIPB), that is, normalized IPB:

- The frequency of blocks with errors (CBSO), that is, the ratio of the number of blocks with errors to the total number of transmitted blocks.

- Frame Information (IR):

IR is the information of the received decoded bit in one coding unit, and it can be interpreted as quantized throughput, that is, the number of correctly received bits per coding unit.

- Block frequency without errors (BWB) , i.e. normalized frame information:

- Throughput (PS)

_bitinfo where R is the transmission rate of information bits _blokkodirov and T is the period of one coding block.

Mutual information-based quality indicators such as IPB, CHIPB, IK and BWBO represent the quality of service and can be used to express transmission requirements / restrictions, for example, IK _target or IIS _target . By comparing the target value with the corresponding measured values, for example, IR _measurement or IPB _measurement , it can be determined whether the requirement is met. Indicators based on mutual information are independent of the sample and channel changes, which makes them more convenient than conventional QoS parameters for use in resource allocation.

It should be noted that other quality indicators based on mutual information, such as, for example, other types of normalized IR and IPB parameters, are also included in the scope of the present invention.

Mutual Information Modulation Model

The modulation model (12 in FIG. 1) deals with mutual SI information at the symbol level for various modulation groups.

According to Shannon’s information theory [5], the channel capacity for a channel with Additive White Gaussian Noise without bandwidth limitation is expressed as follows:

In digital modulation, the SI mutual information indicates the bandwidth of a channel with a digital input and a continuous output. The bandwidth of the group in order of M cannot be higher than log2M, but it can be quite close to the Shannon bandwidth at very low SIR values in the case of excellent knowledge of γ _j . In addition, for a given state γ _{j of the} channel, SI is greater for higher order modulation in the case of excellent channel knowledge. However, it can be expected that in the case of an unsatisfactory channel estimate, the information content will be limited to the estimate γ _j .

Mutual Information Coding Model

As shown in FIG. 1, the coding model 13 for a multi-state channel includes a function for combining 14 and display 15 the quality of the symbol information. The quality model 100 for a channel with many states can, for example, be implemented through look-up tables and include the following steps:

[1] For a set of program outputs of a demodulator with many states { γ ₁ , γ ₂ , ..., γ _J }, the channels { SI ₁ , SI ₂ , ..., SI _J } are calculated by checking the mutual information lookup table for a particular group, as described by the aforementioned modulation model 12.

[2] Select lookup tables for the codec. Tables are generated based on the results of modeling the Additive White Gaussian Noise, which should not depend on the modulation mode. For example, lookup tables for translating IPB to IR or CHIPB to BSEO can be selected.

[3] Assemble the IPB or CHIPB using formula (7) or (8). This function in Fig. 1 is performed in node 14. To model the codec characteristic in the case of a non-optimal coding algorithm, a modification in formula (7) is necessary by making adjustments for the IPB in the form of the IPB correction factor of the IPB _code , in the following form:

Another adjustment method for RBI is expressed as follows:

where the γ _code is the correction factor in the area of the SCE.

[4] Obtain quality indicators by checking ABGS reference tables. This function is shown in figure 1 in node 15.

Representation of a parameter based on mutual information at the block level, such as an IPB (or CHIPB), in the connection-system interface allows separate modulation and coding models, respectively, and the interface between the modulation model 12 and the coding model 13 is linear. A functional feature of the linear interface provides the ability to relatively easily access estimates of various quality indicators based on measurements of connection information.

The above-described quality model proposed according to the present invention is associated with the advantage that it is more accurate than the corresponding connection-system interfaces of the prior art.

New resource allocation procedure

According to the present invention, it is proposed to use rich feedback of the above type, which carries channel state information and transmission information requirements in order to achieve an improved resource allocation procedure. The resource allocation preferably comprises power distribution ( power control) and / or link adaptation . Mainly VI-based indicators of the quality of the previous (current) transmission, prediction of the channel of the subsequent transmission (“next attempt” in the case of using the invention to improve retransmissions) and usually also one or more quality requirements are used to determine the amount of resources, including time resources, frequencies and the power that must be used for subsequent transmission ("second attempt").

Thus, the present invention provides new quality measures in the allocation of resources when, as the distribution of resources according to the prior art, is based on conventional measurements, for example, OSB or BSES. As will be apparent in the future, new indicators based on the VI are associated with some advantageous functional features that provide the possibility of more reliable and efficient packet transmission through communication connections.

For the purpose of disclosure , resource allocation refers to the allocation / assignment / installation / management of resources, such as power resources or resources defining a connection (e.g., related to frequency or time), to a communication connection. The allocation of connection-determining or connection-related resources includes, for example, adaptive selection of a modulation mode, coding rate, and / or source data rate based on channel conditions. It also includes channel allocation in the frequency domain, in the time domain, in the space domain, and / or in the code domain. The distribution of connection-defining resources will then, in general, be referred to as adaptation of the connection.

2 illustrates a (sub) communication system 200 in which the present invention can be used. A base station 21 is illustrated that communicates with mobile nodes 22 (user equipment, mobile stations, and the like) through respective communication connections 23. The invention is particularly useful for (although not limited to) packet communication over wireless connections and is intended for situations where a transmitting node transmits or attempts to transmit a signal to a receiving node through a communication connection. The receiving node monitors the connection, and based on the measured connection information, it is determined how to allocate resources to the connection. In general, all participating nodes 21, 22 are transceivers containing both a receiving function and a transmitting function. For the purpose of the invention, the “receiving” or “transmitting” node may be a base station 21, a mobile node 22, or any other suitable communication node / device.

FIG. 3 is a flow chart of a method for allocating resources according to an embodiment of the present invention. In the first step S1, the signal is transmitted from the transmitting node to the receiving node through a communication connection. The current value of the quality indicator / measurement of the connection (for example, IR, IPB) for the signal is determined based on the relationship / formula of mutual information (step S2). For this purpose, a quality model based on information can be used that simplifies the selection of the modulation mode and adaptation of the coding rate by allowing the selection of the modulation mode and adaptation of the coding rate individually. The quality indicator is a parameter (directly or indirectly) representing the mutual information of the signal at the block level. For example, it can represent common encoded bit information of a received block, such as IPB or CHIPB, or represent common decoded bit information of a received block, such as IR or BWB. The connection information (eg, SIR) of the signal, measured at the receiving node, is preferably used as input in the step of determining the quality indicator.

The step of determining a quality indicator preferably comprises modeling the parameters of the mutual information at the symbol level, using a modulation model with the representation of the signal / noise ratio as input and combining the parameters of the mutual information into mutual information at the block level. This encoded mutual information at the block level can be converted into a decoded quality indicator at the block level using an encoding model independent of the modulation model. For example, step S2 may include mapping the SIR of the received modulation symbol into symbol information (SI); mapping the value of SI in the information of the received block (IPB); and mapping the IPB value to the frequency of the errored blocks (BSEC) and / or frame information (IR).

Also referring to FIG. 3, in step S4, a decision is made on how resources for the communication connection should be allocated in response to the current value of the quality indicator. A quality indicator is typically introduced into a resource allocation function, and resources are then allocated based on the result of the function. Typically, the allocation of resources includes and / or follows a comparison of the current value of the quality indicator and its target value (step S3).

Thus, resource allocation can be performed by a resource allocation parameter, such as power or coding rate. Performing an actual allocation of resources typically involves setting one or more of the following parameters in the sending node:

i) transmission bandwidth and spectral position

ii) duration of transmission

iii) transmit power

iv) packet or subpacket formats in a hybrid automatic repeat request (HARP) session

v) the number of retransmissions in the HAPP session

When a resource allocation is used to determine the transmission bandwidth and its spectral position (i), it, for example, includes setting the position and number of transmitted subcarriers in a multi-carrier system and / or the number of code channels in multiplexed code division systems. When the resource allocation is used to determine the duration of the transmission (ii), it, for example, contains the establishment of the time of transmission of the packet or subpacket in the HARQ session and the duration of the transmitted packet or subpacket in the HARQ system. When a resource allocation is used to determine the formats of a packet or a plurality of subpackages in a HARQ (iv) session, for example, it contains a selection of one or more of the following parameters: modulation order, forward error correction rate (FEC), FEC code type, and HARQ combination type.

Through the invention, it often seems possible to use a single lookup table for a particular coding scheme that is independent of the modulation mode. This leads to a simple implementation of the resource allocation function.

It should be noted that a resource allocation decision can sometimes imply that a transfer should not take place, that is, no resources are allocated to the connection. This is usually regulated by the duration of the transmission; if the current state of the channel or its state in the near future is so bad that the (retransmission) transmission is judged to be useless, the transmission may be paused and resumed later.

As for the criteria for QA, as a rule, at least one criterion of QA that is directly related to the criterion of a quality indicator (for example, IPA _goal or IR _goal ) should be included in determining the distribution of resources. This is usually the case with the criterion for the frequency of blocks with errors or the criterion of bandwidth per connection. However, optional criteria may also be present, such as BSE, packet transfer delay, delay time fluctuations, or residual error block frequency (BSE), which can be used to prioritize individual user services. In any case, when at least one QoS criterion uses a quality indicator similar to IPB or IR, the invention improves the accuracy of the resource allocation function.

The present invention leads to several advantages. It can improve transmission reliability by allocating resources based on channel conditions and quality requirements. Moreover, in the event of a transmission failure, the proposed resource allocation will increase the likelihood of a successful retransmission. This means that there will not be as many retransmissions as with conventional technology, that is, the transmission delay caused by incorrect transmission is reduced.

In addition, through the invention, transmission efficiency can be improved. Adaptation of the connection and power control based on mutual information provide the ability to allocate appropriate resources for retransmission, and accordingly, the useless waste of resources can be reduced. The resource is distributed in an amount slightly larger than sufficient to maintain the quality requirement. Even though a resource allocation threshold is required to ensure stability against channel changes and estimation errors, the useless waste of resources will be less than, for example, with conventional AZP / HAZP systems that support the same quality and transmission delay.

In particular, by introducing power control and / or adapting the connection based on rich feedback, improvements in transmission / retransmission mechanisms such as AZP / HARP can be achieved. For example, through the proposed adaptation of the connection, a flexible choice of coding rate can be provided, that is, variable sizes of retransmission blocks can be provided, and accordingly, the transmission efficiency will be increased in comparison with, for example, the existing GAZP / AMK system in VNRK (Normal AZP / GAZP with PP / NPP feedback can provide only specific types of coding rates, therefore, sometimes it is impossible to avoid wasting a certain amount of resources, even with AMC planning).

There may be embodiments of the invention in which resource allocation includes either power allocation or link adaptation. However, the invention also provides the possibility of implementing a particular preferred embodiment with joint power distribution and adaptation of the connection. Through the above-described information-based quality model, link adaptation and power distribution can be performed simultaneously on the basis of the same measurement (i.e. based on the mutual information of the quality indicator) to achieve the QA requirement with greater accuracy. Such joint power distribution and link adaptation are developed jointly, taking into account common system resources. Typically, in the case of limited channel resources, a more flexible power distribution will be used, and in the case of severely limiting transmission power or interference level, a more flexible connection adaptation will be used. It has been shown that such “collaboration” is superior to conventional independent power distribution and link adaptation.

It should be noted that, although there are systems in the prior art that perform both link adaptation and power distribution, in these systems link adaptation and power distribution are independently designed and do not work together. For example, the WCDMA AMR system has interval-based inner ring power control based on an estimate of the SIR at the interval level; external ring power control taking into account VIP, based on BSBS; and adaptation of the connection taking into account the VIP, based on the assessment of the SIR at the level of VIP. Another example is the High Speed Downlink Packet Access (GNI) system, which uses a connection adaptation tailored to the VIP according to channel status and QoS requirements, along with rather slow power control that does not aim at any QoS requirements.

New measurements and procedures are accordingly very useful in adapting the connection and power distribution for a given user. As mentioned, they are also useful for channel distribution or scheduling where channel resources are used among various users. In other words, for the purposes of the present disclosure, the term “communication connection” can mean both a subconnection in a group of subconnections forming connections / channels to a specific user, and to corresponding connections / channels associated with corresponding users. In the frequency domain, channel allocation refers to carrier allocation (frequency division multiple access (FDMA)) or signal subcarrier (orthogonal frequency division multiplexing (OFDM / OFDMA)). In the time domain, it refers to the allocation of time intervals (Time Division Multiple Access (TDMA)). Similarly, in space, channel allocation refers to the distribution of antenna lines (for example, a system with many inputs and many outputs (MIMO)), while in the code area it refers to the distribution of shared codes (Code Division Multiple Access (CDMA)).

As mentioned in the Prior Art section, conventional HAZP / AMC solutions typically use the required average OSB value as an indicator in the task of counteracting a change in the wireless channel. The usual method relies on the average OSB-BSO ratio. If the current transmission does not reach the desired BSBS, the strategy is to retransmit with such a power level or with such another modulation or coding format that the total received SIR will be sufficient for the desired BSBS. However, the average OSP-CHBSO characteristic depends on the rate and image of the channel change. Different rates and images of the channel change lead to different curves of the average OSP-CBSO ratio even with the same modulation and coding. GAZP / AMK require exhaustive modeling of all possible channel changes, which is a difficult task. The invention overcomes these difficulties by presenting a measurement characteristic of information (quality indicators based on VI) that interprets the changeable values of the SCE in a uniform amount, which is independent of the speed and image of the channel change.

4A-4C are block diagrams illustrating various ways of organizing a quality indicator determination function in communication nodes according to embodiments of the present invention. All illustrated resource allocation mechanisms 400 represent a receiving node 41 with a line monitoring function 43 and a transmitting node 42 with a function 45 for performing actual resource allocation.

In the first embodiment (FIG. 4A), a quality indicator is calculated in the quality indicator calculator 44 of the receiving node 41. The receiver 41 also comprises means (not shown) for deciding which resources should be allocated to this connection. A resource allocation control command is sent from the receiving node 41 to the transmitting node 42, and the transmitting unit allocates resources according to this command.

In a second embodiment (Fig. 4B), the quality indicator calculator 44 is instead located on the side of the transmitter. The receiver 41 simply transmits the connection information for calculating the quality indicator to the transmitter 42, after which the quality indicator is calculated in the transmitter, which determines and performs the allocation of resources.

As illustrated in FIG. 4C, there may also be embodiments where the quality indicator (s) are calculated (s) in a location that is neither a live receiver nor a transmitting node, for example, in a separate top-level control node 46. In this case, the connection information for calculating the quality indicator is transmitted from the receiver 41 to the external control unit 46. The control unit 46 calculates a quality indicator and transmits a resource allocation control command to the transmitter 42. As in FIG. 4A, the resource allocation function 45 of the transmitter 42 then performs resource allocation according to the control command.

In order to avoid confusion, in FIG. 4, nodes 41 and 42 are designated as receiver and transmitter. Of course, in a normal situation, each node contains both a receive function and a transmit function, that is, it is a transceiver node.

Quality Indicator Definition - Implementation Example

As mentioned, the invention has several useful applications in conjunction with retransmissions of the AZP / HAZP.

In HARQ, the frequency of the source is fixed, that is, equal to the constant K, and, as mentioned in the section "Background", the software combination procedure depends on which HAZP combining scheme is used. With GAZP-PC, the receiver always combines the entire retransmission of the block with errors; that is, the amount of data that the receiver buffers remains unchanged. In HARQ-II, the receiver buffers encoded characters that represent new information to the block transmitted first; that is, the amount of data that needs to be buffered increases with subsequent retransmissions.

Accordingly, there are three types of AZP / GAZP schemes: ordinary AZP (type I); HAZP with Direct Combination (type II); and GASP with Incremental Redundancy (Type III). The allocation of resources in various types can be explained through various factors that are regulated / determined:

• For types I and II, that is, for ordinary AZP and GAZP-PK, since the encoding rate is fixed, the BSBO _target and IR _target can be converted into an IPB _target. Therefore, adaptation of the retransmission will be to minimize power consumption by adjusting power to achieve

IPB _goal , that is, in the distribution of power . Typically, the goal is to determine the power needed to achieve the adopted OSB, which most closely matches the IPB _target , using GAZP-PK.

• For type III, that is, GAZP-II, the goal, as a rule, will be to minimize the channel’s resources and achieve the IR _target , that is, the adaptation of the coding rate . The coding rate varies according to the number of retransmissions and strategies. After each retransmission, the HARQ-AI system will match the specific pattern code, SPI _{purpose of} which can be easily obtained from the _{target IR.} It also seems possible to combine coding rate adaptation with power distribution. Thus, the goal may be to determine the coding rate necessary to achieve the adopted IPB value, which is the closest _target to the IPB using GAZP-NI.

In addition, for all cases, modulation adaptation can be performed based on the instantaneous channel quality. Modulation adaptation is usually selective, in the sense that manufacturers may choose not to implement modulation adaptation.

For an encoding block with K bits of information, equation (15) provides frame information (IR) for the i-th transmission with the following expression:

According to the target BSES (or other requirements of the QA), the requirements of the IR are derived as:

For a particular encoding mode, there is an exceptional correspondence between the IR and the IPB information of the encoded (received) block. IPB for the i-th received block is designated as IPB _i . Accordingly, target IPB values for various coding modes can be obtained.

Based on the measurements of information of failed i times receptions, the difference between the purpose of the information of successful reception can be calculated and sent back to the transmitter as a requirement for information for the next retransmission. For AZP, the information requirement for the i + 1st retransmission attempt is defined as:

For HAZP, it is expressed as the difference in information:

Or in the field of IPB:

Moreover, for a given modulation mode, a correspondence between ΔIPB and ΔOSP can be constructed.

Typically, IR expressions are used to select a coding mode, while IPB expressions are used to select a modulation mode and power control.

The following paragraphs provide examples of power control, modulation mode selection, and coding rate selection for the AZP / GAZP schemes.

Power control and modulation mode selection for a given coding mode

For a given coding mode, the ΔIK requirement of the frame information can be calculated. For both the AZP scheme and the GAZP-II scheme, the IPB requirement for the next retransmission attempt is calculated as:

with relevant OSB requirement:

where ΔN _{i + 1} is the number of coded bits in the coding block for the upcoming (i + 1) -th retransmission, which is N for the AZP and GAZP-PC. The transmitted power of the (i + 1) -th retransmission must be determined by Δ OSP _{i + 1} and channel prediction.

For GAZP-PC with combination with the maximum ratio (KMO), the requirement of the SCE for the next retransmission attempt is defined as:

with the relevant requirement of the IPB:

The SIR requirement is an effective SIR target for power control in the next retransmission attempt, that is, the power must be distributed so as to achieve ΔIPB with a given modulation mode.

5 contains a graph representing an example of a power distribution for a HAZP-PC system. The mapping function between IPB and OSB is shown.

OSB _{E, 1 ~ t-1} is the effective combined SCE of the previous 1 ~ (t-1) gears, and OSP _{E, 1 ~ t} is the desired effective combined SCE for all t gears. OSB _{E, t} is the target OSB of the t-th transmission, on the basis of which a power distribution decision is made. IPA _goal gives the requirement for the final combined OSB _goal. Combined SCE can be obtained based on measurements of all received transmissions. The differences between the SIR _target and measurements, together with the prediction of channel quality, give rise to a power requirement.

As for the modulation mode, for a given effective target SIR, which can be limited by the maximum power threshold, the modulation mode should be chosen so as to satisfy the ΔIPB requirement for the upcoming (retransmitted) block.

The present invention provides the ability to adapt modulation in a coding unit to maximize channel throughput. This is another of its advantages. By presenting a VI-based quality model, a plurality of modulation modes can be selected in a coding unit, and using an appropriately designed algorithm, a mixed modulation scheme can be better than a single modulation scheme.

Coding rate selection

The choice of coding rate is mainly used with GAZP-II.

In the case of limited transmit power and a given modulation scheme, the coding rate for the next retransmission attempt should be determined so as to satisfy the ΔIK requirement.

Let R of _losses denotes the ratio of losses on the path of the forecast channel of the i + 1st channel transmission in the i-th unsuccessful attempt, then:

where P _i and P _{i + 1} are the transmitted powers for the i-th and i + 1-th retransmissions, respectively, and SIR _i and SIR _{i + 1} are the corresponding values of the received SIR. Accordingly, the average symbolic information (SI) can be calculated as follows:

After the (i + 1) -th attempt of retransmission to the HAZP-II, the coding rate will change from (N _i , K) to (N _{i + 1} , K), i.e., the (i + 1) th attempt will transmit ΔN = N _{i +1} -N _i encoded bits. The mapping function of the ICBIP _{code, i + 1} () for the (i + 1) -th attempt is determined by the coding rate. The coding rate should be chosen so as to satisfy the following condition:

6 contains a graph illustrating the choice of coding rate for the HARQ-II system. Several mapping functions between IR and IPB are shown for various coding rates Δ. The information requirement (IPB requirement) of the next transmission for each coding rate can be calculated from

IR _target and measured adopted by IPB. Based on the channel prediction and the corresponding power distribution, the next transmission IPB can be predicted. The highest coding rate that satisfies the IR _target will be selected. In the illustrated example, this means that Δ1 / 2 will be selected. Then N is determined by the final speed of the t-th and (t-1) -th gears. Alternatively, for a selected coding rate, the power can be determined using the IPB requirement and channel quality.

7 shows a block diagram of a system for allocating resources from a HARQ-PC according to an embodiment of the present invention.

An example illustrates a system / mechanism 700 for combined power distribution and link adaptation, comprising: a quality display unit 70; node 71 (retransmission) transmission; channel 72; receiver 73; a node 74 for channel prediction; node 75, 76 for combining the SCE and calculating the IPB; node 77 for comparing the IPB; a resource allocation decision unit 78; and a node 79 for controlling (distributing) power and adapting the connection.

The quality display unit 70 is provided with or has access to a number of coding characteristic lookup tables (e.g., IPB-CHBSO or CHIPB-CHBSO for a number of coding modes). The inputs to the quality display unit 70 include at least one quality requirement, such as a BSEO _target , and encoding mode information, such as encoding rate and block size. Through these inputs, the quality display unit can check the information requirement / quality indicator for each coding unit. In Fig.7, the target value of the quality indicator IPB-IPB, the _target is displayed from the node 70 display quality.

Turning to the function of a transceiver containing a connection in which channel measurements are made, a sequence of information bits is input to the transmitter unit 71. The transmitter unit 71 typically has means for a first transmission as well as for retransmissions, and includes a coding / modulation / buffering sub-unit 71A and a power distribution sub-unit 71B. The coding / modulation / buffering subunit 71A receives inputs containing a sequence of information bits and QMS signals (modulation mode, coding rate, etc.) and outputs a modulated symbol sequence. The modulated symbol sequence is redirected to the power distribution unit 71B, to which the current transmitted power is also supplied. The node 71B provides a transmitted symbol sequence with a distributed power level.

Before entering the receiver unit 73 on the receiving side, the transmitted main-sequence symbol sequence passes through the channel. For a real system, the channel assembly 72 of FIG. 7 represents a radio transmission in air, while for modeling purposes it can be a channel model presented for modeling some typical wireless propagation channels.

At the receiver unit 73, the received baseband symbol sequence is subjected to measurement and evaluation functions that are used to perform link adaptation and power control. The findings from the receiver unit 73 include channel impulse response estimates that are transmitted to the channel prediction unit 74, and information related to the SIR information, such as the SIR estimates for each transmission node (eg, time intervals) or the distribution of SIRs, which are typically transmitted to a channel prediction unit 74, as well as an SCE combining unit 75. The channel prediction unit 74 predicts the distribution of the SCE of upcoming transmissions based on the estimates / distributions of the SCE (for each transmission node) of the previous (several previous) VIPs and the channel impulse response of the previous VIPs.

Moving to the SCE combining unit 75, the combination with the maximum ratio (KMO) is accepted, and inputs containing SCE estimates for each transmission node of the previous (several) VIPs are combined into an effective SCE (SCO _i ) after combining the KMO. Further, in the node 76 computing the IPB, the equivalent IPB for the total number i of transmission attempts is calculated based on the equivalent SCE of the total number of 1 ~ i transmissions and the modulation mode.

System 700 further comprises a function for comparing a target value of a quality indicator (IPB) with its measured / estimated value. This function in Fig. 7 is represented by the IPB comparison unit 77, which basically performs the following calculation: ΔIPB = IPB _target - IPB _i . The comparison result is translated into the solution ΔIPB in the node 78 of the solution ΔIPB. In this example, if ΔIPB≤0, then the allocation of resources will not be further increased, but, on the other hand, if ΔIPB> 0, then the IPB requirement of the next transmission (for example, expressed through ΔIPB) is transmitted to the power control / adaptation unit 79. In this node 79 for resource allocation, thus, the IPB requirement for the next transmission and the statistical forecast of the SCE for the next transmission are provided. Based on these inputs, he can perform QMS selection and power distribution so that transmission conditions improve.

Typically, resource allocation, like the mechanism described above, is an iterative process where the outputs of the power distribution / adaptation unit 79 are sent back to the quality display unit 70 and the transmitter unit, respectively, and continuous measurements are taken on the received symbol sequence in order to improve resource allocation and installation transmission.

FIG. 8 is a structural diagram of a system for allocating resources with HARQ-II according to an embodiment of the present invention. An example illustrates a system / mechanism 800 for combined power distribution and link adaptation, comprising (s) a quality display unit 80; node (retransmission) transmission; channel 82; receiver 83; a node 84 for channel prediction; nodes 85 for calculating the IPB; node 87 for comparing IPB; node 88 decision ΔIPB and node 89 for control (distribution) of power and adaptation of the connection.

Most of the functions of FIG. 8 correspond, more or less accurately, to the functions described with reference to FIG. However, FIG. 8 illustrates HARQ-II, while FIG. 7 illustrates HARQ-PC. This means that in FIG. 7, the coding rate is fixed for all retransmissions, that is, the quality correspondence (IPB-CHBSO) for HAZP-PC is fixed. Therefore, the IPB _target needs to be checked only once. However, in FIG. 8, the coding rate decreases with an increase in the number of retransmission attempts due to the increasing number of redundant bits. Consequently, IPB-CBSO compliance varies depending on the number of retransmissions and strategies. Therefore, the IPB _target must be checked for each retransmission attempt.

Moreover, the findings of nodes 73; 83 receivers vary. In the case of GAZP-PC, an assessment of the SIR and channel estimation are required. GAZP-II uses another additional assessment - the assessment of the quality indicator (IPB).

Regarding the choice of QMS in node 71; 81, with the GAZP-PC system, in all retransmission attempts, only a specific QMS is used, therefore, for a particular block of information, encoding and modulation is performed only once, and the buffered modulated symbol sequence will be distributed at different power levels, i.e., it will be controlled by the power distribution in all retransmissions. On the other hand, with the GAZP-II system, both the encoding rate and the modulation mode will be selected for any retransmission attempt, as well as the power distribution.

With modifications obvious to those skilled in the art, the circuits illustrated in FIGS. 7 and 8 can be used to perform a separate power distribution or a separate adaptation of the connection. IPB can be replaced by another quality indicator based on mutual information, such as the above-described parameters at the block level. Moreover, it should be understood that the nodes of FIGS. 7 and 8 represent functions that are preferably represented in the resource allocation system according to the invention. Various embodiments may have functions otherwise implemented, and two or more nodes may be successfully implemented together in one physical device.

Taking into account the coding rate limitation associated with the HARQ-II, i.e., a lower combining gain for a lower coding rate, joint power distribution will be very useful in this case. (The problem is serious enough for turbo mode, since most existing systems use 1/3 of the turbo coding rate as the main code, and these systems use speed comparison to control the resulting coding rate. GAZP-II is known to have a fairly small advantage compared to GAZP-PC for a coding rate of less than 0.5.)

The above models for AZP / GAZP-PK and GAZP-II are useful for reducing retransmission delay.

From the above examples, it is obvious that the procedure according to the invention is very useful in cases where there was a transmission error and retransmission is performed. However, it should be emphasized that the invention is equally applicable to provide a decision on what resources are required to transmit a new packet. For example, for connections without HARQ (such as live voice), rich feedback can be used to control the power of the next packet to maintain the desired level of efficiency. Even when the current packet is received successfully, rich feedback can tell the system if there are enough resources to maintain the desired level of performance for future packets. In a particular preferred embodiment of the invention, comparisons with target values and adjustments to distributed resources are performed continuously to increase transmission reliability and efficiency.

Although the invention has been described with reference to the specific illustrated embodiments, it should be noted that it also encompasses equivalents of the disclosed functions, as well as modifications and variations obvious to those skilled in the art. Therefore, the scope of the invention is determined solely by the attached claims.

ABBREVIATIONS

ABGSh - Additive White Gaussian Noise

AZP - Automatic Repeat Request

AMK - Adaptive Modulation and Coding

VI - Time Interval

VIP - Transmission Time Interval

GNI - High Speed Downlink Packet Access

VNRK - High Speed Downlink Shared Channel

GAZP - Hybrid Automatic Repeat Request

GAZP-NI - GAZP with Incremental Redundancy

GAZP-PK - GAZP with Direct Combination

DFM - Binary Phase Manipulation

IR - Frame Information

IPB - Adopted Block Information

IPI - Accepted Interval Information

KMO - Combination with Maximum Ratio

KO - Quality of Service

CPM - Quadrature Phase Manipulation

NPP - Non-Confirmation of Admission

OSB - Signal to Noise Ratio

FFP - Direct Error Correction

PP - Admission Confirmation

SI - Modulated Character Information

QMS - Modulation and Coding Scheme

BWC - Block Frequency Error Free

BSBS - Error Block Frequency

CHIPB - Received Block Information Frequency

EOE SCE - Exponential Mapping of Effective SCE

BIBLIOGRAPHY

Claims (40)

1. A method for allocating resources for communication connections, comprising the steps of:
determine the current value of the quality indicator for the signal transmitted through the connection (23) communication from the transmitting node (21, 22; 42; 71; 81) to the receiving node (21, 22; 41; 73; 83) as a value based on mutual information ; and
decide on the allocation of resources to connect the connection in response to the current value of the quality indicator. 2. The method according to claim 1, in which the quality indicator represents the mutual information of the signal at the block level. 3. The method according to claim 1 or 2, in which the information (11) of the signal connection, measured at the receiving node (21, 22; 41; 73; 83) is used as input information at the stage of determining the quality indicator. 4. The method according to claim 1, in which the quality indicator represents the total information of the encoded bits of the received block. 5. The method according to claim 1, in which the quality indicator represents the total information of the decoded bits of the received block. 6. The method according to claim 1, in which the step of determining a quality indicator, in turn, comprises the steps of:
model parameters of mutual information at a symbolic level using a modulation model (12) with presentation of signal / noise ratio (11) as input information; and
combine the parameters of mutual information into mutual information at the block level. 7. The method according to claim 6, in which the step of determining the quality indicator also comprises the step of determining a quality indicator at the block level based on the coding model (13) independent of the modulation model (12), and the step of using combined mutual information at the block level. 8. The method according to claim 1, in which the step of deciding on the allocation of resources, in turn, comprises the steps of:
comparing the current value of the quality indicator with the target value of the quality indicator; and
allocate resources based on the difference between the current and target values of the quality indicator. 9. The method according to claim 1, in which the step of deciding on the allocation of resources is also based on at least one quality of service requirement selected from the group which includes the frequency of error blocks (BSEC), packet transmission delay, time fluctuation delays, residual error bit rate, and service priority. 10. The method according to claim 1, in which the distribution of resources comprises the distribution of power on the connection (23) connection. 11. The method of claim 10, which is used with a Hybrid Automatic Repeat reQuest (HARQ) with direct combining and comprises a power distribution to achieve the target value of the total information of the encoded bits of the received block. 12. The method according to claim 1, in which the distribution of resources comprises adapting the connection (23) connection with respect to at least one parameter defining the connection. 13. The method according to claim 1, wherein the resource allocation comprises joint connection adaptation and power distribution. 14. The method according to item 12 or 13, which comprises adapting a connection-determining parameter selected from the group consisting of the coding rate, coding mode, and modulation mode. 15. The method according to 14, which is used with HARQ with incremental redundancy and comprises adapting the coding rate to achieve the target value of the representation of the total information of the encoded bits of the received block. 16. The method according to 14, which contains the adaptation of the modulation mode and the adaptation of the coding rate, performed separately. 17. The method according to item 12 or 13, in which the adaptation of the connection, contains the distribution of channels to appropriate users. 18. The method according to claim 1, containing stages in which:
calculating a quality indicator in the receiving node (41); and
transmitting a resource allocation control command corresponding to the decision to allocate resources from the receiving node to the transmitting node (42), whereby resources can be distributed in the transmitting node according to the control command. 19. The method according to claim 1, containing stages in which:
in the transmitting node (42) receive connection information from the receiving node (41) to calculate the quality indicator;
in the transmitting node, a quality indicator is calculated; and
by the transmitting node, resources are allocated in the transmitting node based on the resource allocation decision. 20. The method according to claim 1, containing stages in which:
in the external control node (46), receive connection information from the receiving node (41) to calculate a quality indicator;
calculate the quality indicator in the external control node; and
transmitting the resource allocation control command corresponding to the resource allocation decision made from the external control node to the transmitting node (42), whereby the resources can be distributed in the transmitting node according to the control command. 21. A communication node (21, 22; 41, 42) in a system (200; 400; 700; 800) with means for allocating resources for communication connections (23), comprising:
means (44) for determining a current value of a quality indicator for a signal transmitted through a communication connection as a value based on mutual information; and
means for deciding on the allocation of resources to connect the communication in response to the current value of the quality indicator. 22. The communication node according to item 21, in which the quality indicator represents the mutual information of the signal at the block level. 23. The communication node according to item 21 or 22, in which the measured information (11) of the signal connection is used as input in the definition of a quality indicator. 24. The communication node according to item 21, in which the quality indicator represents the total information of the encoded bits of the received block. 25. The communication node according to item 21, in which the quality indicator represents the total information of the decoded bits of the received block. 26. The communication node according to item 21, in which the means for determining the quality indicator, in turn, contains:
means for modeling parameters of mutual information at a symbolic level, which includes a modulation model (12) with the representation of the signal / noise ratio (11) as input information; and
means for combining parameters of mutual information into mutual information at the block level. 27. The communication node according to claim 26, wherein the means for determining at least one quality indicator also comprises means for determining a quality indicator at a block level based on a coding model (13) independent of said modulation model (12), and to use combined mutual information at the block level. 28. The communication node according to item 21, in which the means for deciding the allocation of resources, in turn, contains:
means for comparing the current value of the quality indicator with the target value of the quality indicator; and
means for allocating resources based on the difference between the current value and the target value of the quality indicator. 29. The communication node according to item 21, in which the means for deciding on the allocation of resources uses at least one quality of service requirement selected from the group which includes the frequency of blocks with errors (BSEC), transmission delay, fluctuation of the delay time, residual error bit rate and service priority. 30. The communication node according to item 21, in which the distribution of resources contains a distribution of power for connection (23) communication. 31. The communication node according to claim 30, which is used with a Hybrid Automatic Repeat Request (HARQ) with direct combination and comprises means for allocating resources to achieve a target value for representing the total information of the encoded bits of the received block. 32. The communication node according to item 21, in which the distribution of resources comprises adapting the connection (23) of the connection with respect to at least one parameter defining the connection. 33. The communication node according to item 21, in which the distribution of resources includes joint adaptation of the connection and the distribution of power. 34. The communication node according to claim 32 or 33, which comprises adapting a connection-determining parameter selected from the group that includes the coding rate, coding mode, and modulation mode. 35. The communication node according to clause 34, adapted to work with HARQ with incremental redundancy and containing the adaptation of the coding rate to achieve the target value of the representation of the total information of the encoded bits of the received block. 36. The communication node according to clause 34, which comprises means for separately adapting the modulation mode and adapting the coding rate. 37. The communication node according to item 21, which also contains:
means for receiving a signal from a transmitting node (42) through a communication connection; and
means for transmitting a resource allocation control command corresponding to the decision to allocate resources to the transmitting node, whereby resources can be distributed in the transmitting node according to the control command. 38. The communication node according to item 21, which also contains:
means for transmitting a signal to a receiving node (41) through a communication connection;
means (44) for determining the current value of the quality indicator using the connection information from the receiving node; and
means (45) for the allocation of resources according to the decision on the allocation of resources. 39. The communication node according to item 21, which also contains:
means (44) for determining the current value of the quality indicator for the signal transmitted through the communication connection from the transmitting node (42) to the receiving node (41) using the connection information from the receiving node; and
means for transmitting a resource allocation control command corresponding to the decision to allocate resources to the transmitting node, whereby resources can be distributed in the transmitting node according to the control command. 40. A communication system with means for allocating resources for communication connections, the system comprising a communication node according to any one of paragraphs.21-39.

Images