JP5815574B2 - Bandwidth allocation method and bandwidth allocation apparatus - Google Patents

Description

  The present invention relates to a bandwidth allocation method and a bandwidth allocation device that divides a bandwidth of an optical transmission system into slots along a time axis and an optical frequency axis and allocates the slots to communication between a parent node and a child node.

  Research and development for transmitting and receiving OFDM (Orthogonal Frequency Division Multiplexing) signals on an optical transmission system typified by a PON (Passive Optical Network) system has been actively conducted. An OFDM signal can be generated by IFFT (Inverse Fast Fourier Transform) operation, and can be decoded by FFT (Fast Fourier Transform) operation. Use of OFDM technology along with improvements in signal processing technology and integrated circuit technology Expectations for optical transmission systems have been increasing.

  The PON system is a point-to-multipoint (P2MP) type optical transmission composed of one OLT (Optical Line Terminal) corresponding to a parent node, a plurality of ONUs (Optical Network Units) corresponding to child nodes, and an optical transmission cable connecting them. System. In the PON system using the OFDM technology, the OLT and the optical transmission cable are shared by frequency division multiplexing and time division multiplexing. In the following description, a child node may be expressed as “user”.

  In the OFDM system, data can be assigned to two axes of time and carrier frequency (hereinafter, simply referred to as “carrier wave”). In a wireless system using OFDM, a transmission path environment that changes depending on user position information and time is measured, and time and frequency allocation scheduling is performed so that transmission efficiency is improved as a whole system. On the other hand, in the OFDM-PON system, which is an optical transmission system, the change in the transmission path environment with time is small. Up to now, the order relation between the data addressed to other ONUs to be transmitted / received at the same time and the assigned slot order has not been considered, and scheduling according to the number of slots requested by each ONU has been studied (for example, Non-Patent Document 1). See).

For example, as a time and carrier wave allocation method in an optical transmission system, there is a method (hereinafter referred to as related method 1) in which data is allocated in the time direction as shown in FIG. In the example of FIG. 1, time and carrier wave allocation control are performed in the order of ONU1, ONU2, ONU3, and ONU4. By performing allocation control so that slots in the carrier wave direction are used to the maximum, the time T _idle during which no signal is transmitted or received is maximized.

Alternatively, as shown in FIG. 2, there is a method (hereinafter referred to as related method 2) in which each ONU determines an available carrier in advance and occupies the available carrier. In the related method 2, the time T _idle during which no signal is transmitted / received depends on the ONU having the largest data transmission / reception amount. For this reason, the time T _idle during which no signal is transmitted / received is reduced as compared with the related method 1. On the other hand, since each ONU can occupy a band, it has a feature of achieving high fairness among ONUs.

Japanese Patent Application No. 2011-278686

Jingjing Zhang, Ting Wang, and Nirwan Ansari, “An Efficient MAC Protocol for Asynchronous ONUs in OFDMA PONs”, OFC / NFOEC 2011

  In OFDM-PON, signals are generated and decoded by digital signal processing. In OFDM-PON, adaptive modulation in which a signal generation method such as a symbol modulation / demodulation method is changed according to a user and time is effective. It is possible to provide different bandwidths depending on the user and time by dynamically changing the signal generation method, but the switching frequency of the signal generation method varies depending on the time and the carrier allocation method.

  In addition, the inventors have dynamically changed the calculation accuracy of IFFT calculation and FFT calculation required so far when transmitting and receiving the OFDM signal, thereby reducing the power consumption required for generating and decoding the OFDM signal. A signal generation method employing a dynamic control method has been proposed (see, for example, Patent Document 1). In Patent Document 1, signal generation and decoding are performed using different calculation accuracy depending on the user, and power consumption required for signal generation is reduced by selecting a low calculation accuracy. In Patent Document 1, when users requesting different calculation accuracy are mixed among users who transmit and receive signals at the same time, it is necessary to control the calculation accuracy of FFT and IFFT according to the user who requests the highest calculation accuracy. There is. For this reason, the power consumption reduction effect may change depending on the time and the carrier wave allocation method.

  Both the related method 1 and the related method 2 perform allocation using the amount of downstream traffic and requested bandwidth information from the ONU. Unlike OFDM signal scheduling in a wireless environment, allocation is performed using the number of slots required for data transmission / reception for each ONU. For this reason, what kind of data is contained in other carriers at the same time, how the signal generation method and the signal decoding method change depending on the time, switching frequency and frequency of the signal generation method and the signal decoding method Is not considered. Since such a time and carrier wave allocation method is used, the OFDM-PON system has a problem that it is not possible to achieve optimization of power consumption by frequency modulation switching at the time of adaptive modulation and dynamic control of calculation accuracy.

  Therefore, in order to solve the above-described problems, the present invention provides a bandwidth allocation method and a bandwidth allocation device that reduce the number of times of switching between signal generation methods and improve the power reduction efficiency in the power consumption reduction method according to the calculation accuracy. The purpose is to provide.

  A group is formed according to a signal generation method (symbol modulation / demodulation method or calculation accuracy dynamic control method), slots at the same time are assigned to users belonging to the same group, and users belonging to the same group are consecutive in the time direction. So that it was scheduled.

Specifically, the bandwidth allocation method according to the present invention divides a bandwidth of an optical transmission system in which a parent node and one or more child nodes are connected by an optical transmission path into slots along a time axis and an optical frequency axis, Is a bandwidth allocation method for allocating to a communication with the child node,
A group forming step of grouping the child nodes having the same signal generation method of signals transmitted and received between the parent node and the child notebook, and generating group information of the grouped child nodes;
A slot assigning step of assigning the signals of the child nodes to the slots so as to be continuous in the same time and in the time direction for each group formed in the group forming step;
I do.

  By grouping child nodes into groups and communicating for each group, the number of switching of signal generation methods such as symbol modulation / demodulation methods can be reduced. In addition, since signals having the same calculation accuracy can be transmitted continuously by communicating for each group, signals transmitted with excessive calculation accuracy can be reduced, and power can be reduced. Therefore, the present invention can provide a bandwidth allocation method that reduces the number of times of switching the signal generation method and realizes an improvement in power reduction efficiency in the power consumption reduction method according to the calculation accuracy.

When the signal generation method is a symbol encoding method, in the slot allocation step,
The slot is allocated so as to minimize the switching frequency of the symbol encoding method by referring to the group information of the group formed in the group forming step and the downlink use band information for the child node. The slot allocation step refers to the group information of the group formed in the group formation step and the uplink request bandwidth information from the child node, and sets the slot so that the switching frequency of the symbol encoding method is minimized. Assign and notify the child node.

  When the signal generation method has an operation accuracy required for signal generation and demodulation, the slot allocation step refers to the group information of the group formed in the group formation step and the downlink use band information for the child node. The calculation accuracy switching frequency is minimized, and the number of slots for signal generation with the required calculation accuracy is maximized. The slot allocation step refers to the group information of the group formed in the group formation step and the uplink requested bandwidth information from the child node, minimizes the switching frequency of the calculation accuracy, and demodulates with the required calculation accuracy. Maximize the number of slots to perform.

  The bandwidth allocation method described above is realized by a bandwidth allocation device that includes a storage unit that executes a group formation step and holds the group information, and a signal scheduling unit that executes a slot allocation step, and is arranged in the parent node it can.

  The present invention can provide a bandwidth allocation method and a bandwidth allocation device that can reduce the number of switching of the signal generation method and improve the power reduction efficiency in the power consumption reduction method according to the calculation accuracy.

It is a figure which shows the scheduling which minimizes the frequency | count of execution of FFT and IFFT by performing a carrier wave allocation in order of data arrival. It is a figure which shows the scheduling which implement | achieves high fairness by determining beforehand the carrier wave which can be used with respect to each ONU. It is a figure explaining an optical transmission system provided with the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention. It is a figure explaining the symbol encoding apparatus which the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention controls. It is a figure explaining the bit variable IFFT arithmetic unit which the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention controls. It is a figure which shows the scheduling result which allocated ONU to the same time slot with the bandwidth allocation method which concerns on this invention. It is a figure explaining an optical transmission system provided with the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention. It is a figure explaining the symbol decoding apparatus which the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention controls. It is a figure explaining the bit variable IFFT arithmetic unit which the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention controls. It is a figure explaining an optical transmission system provided with the band allocation apparatus which implement | achieves the band allocation method which concerns on this invention.

  Embodiments of the present invention will be described with reference to the accompanying drawings. The embodiments described below are examples of the present invention, and the present invention is not limited to the following embodiments. In the present specification and drawings, the same reference numerals denote the same components.

(Embodiment 1)
FIG. 10 is a diagram for explaining an optical transmission system 451 including the band allocation device 301 of the present embodiment. The optical transmission system 451 includes a plurality of ONUs 10a as child nodes, an OLT 10b as a parent node, and an optical fiber 10c and a splitter 10d as optical transmission lines. The OLT 10 b includes an optical signal transmission / reception device 106, a band allocation device 301, and a signal generation device 302.

  The optical signal transmission / reception device 106 converts the signal generated by the signal generation device 302 into a downstream optical signal and transmits it to each ONU 10a. Further, the optical signal transmission / reception device 106 receives the upstream optical signal from each ONU 10 a and passes it to the signal generation device 302. Further, the optical signal transmission / reception device 106 receives a bandwidth request from each ONU 10a and passes it to the bandwidth allocation device 301 as requested bandwidth information 108b.

  The signal generation device 302 receives a signal from a higher-level server or the like (not shown), and sets and transmits the effective bit number according to the accuracy of symbol encoding and IFFT operation according to an instruction from the band allocation device 301. Generate a signal. At this time, the downlink use band information 108a is delivered to the band allocation device 301 based on the data held in the data buffer 104a. The signal generator 302 performs symbol decoding and IFFT decoding on the signal from each ONU 10a received by the optical signal transmitter / receiver 106 in accordance with an instruction from the band allocating device 301, and outputs the decoded signal to an upper server or the like.

  The band allocation device 301 executes a group formation step of a band allocation method described later, and includes a storage unit 109 that holds group information described later, and a signal scheduling unit 107 that executes a slot allocation step of a band allocation method described below. And is arranged in the OLT 10b.

The bandwidth allocation method performed by the bandwidth allocation device 301 is a bandwidth allocation method in which the bandwidth of the optical transmission system 451 is divided into slots on the time axis and the optical frequency axis, and the slots are allocated to communication with the ONU 10a.
A group forming step for grouping ONUs 10a having the same signal generation method for signals transmitted and received between the OLT 10b and the ONU 10a, and generating group information 109 of the grouped ONUs 10a;
A slot assignment step of assigning the signals of the ONU 10a to the slots so as to be continuous in the same time and in the time direction for each group formed in the group formation step;
I do.

  The present bandwidth allocation method is characterized by reducing the frequency of switching between the symbol encoding method and the symbol decoding method and improving the power consumption reduction efficiency by controlling the calculation accuracy of IFFT used for OFDM signal generation and FFT used for decoding. Allocation scheduling for time and carrier slots to be performed.

[Downlink scheduling]
FIG. 3 is a diagram for explaining a downlink signal scheduling portion that realizes reduction in the number of symbol coding switching times and improvement in power consumption reduction efficiency by calculation accuracy control in the optical transmission system 451 described in FIG.

Downlink signal scheduling that realizes time and carrier allocation for downlink signals includes data buffer 30, serial-parallel converter 31, symbol encoding device 32, time and carrier allocation device 33, bit variable IFFT arithmetic unit 34, parallel-serial conversion. An OLT 3b having a device 35, an optical signal transmission device 36, a downlink signal scheduling unit 37, downlink use band information 38, and group information 39;
An optical transmission system 451 composed of a plurality of ONUs (3a-1 to 3a-n) that forms a plurality of groups based on the symbol encoding method to be used or the calculation accuracy required for IFFT calculation performed for generating an OFDM signal. In
The downlink signal scheduling unit 37 performs downlink signal allocation control so that data using the same symbol coding scheme is continuous in the time direction, and performs the same downlink signal for IFFT calculation with the same number of effective bit digits. By assigning time and carrier to assign to time slots,
The scheduling is characterized by reducing the switching frequency of the symbol encoding method and improving the effect of reducing the power consumption required for the IFFT calculation.

  Here, the characteristic of the downlink signal scheduling unit 37 is that the ONU 3a having parameters that can change the signal generation method according to the communication partner, such as the symbol encoding method and the number of effective bit digits used in the IFFT calculation. The frequency of switching the signal generation method is reduced by grouping ONUs that use the same signal generation method.

  Further, in the optical transmission system having the bit variable IFFT arithmetic unit 34, the power consumption reduction efficiency of the bit variable IFFT arithmetic unit 34 used for generating the OFDM signal is improved by scheduling to increase the chance of reducing the number of effective bits. It is possible to make it.

(Reduction of symbol coding switching frequency)
In the present embodiment, the signal generation method is a symbol encoding method. In the slot allocating step, the slot is allocated so as to minimize the switching frequency of the symbol coding scheme by referring to the group information of the group formed in the group forming step and the downlink use band information for the ONU 3a.

  FIG. 4 is a diagram for explaining downlink signal scheduling for reducing the switching frequency of the symbol coding scheme for downlink signals. The symbol encoding device 40 includes a data buffer 401, a symbol encoding input rearranger 402, a changeover switch (403, 405), and symbol encoders (404-1 to 404-n) that realize different symbol encoding schemes. Thus, the downlink signal scheduling unit 41 gives scheduling information to the symbol encoded input rearranger 402 and the changeover switches 403 and 405 based on the group information 42.

Here, downlink data addressed to ONU ₁ , ONU _2, ONU _3, ONU ₄ arrives at the symbol encoding device 40 in sequence, ONU ₁ , ONU ₄ is symbol encoded by Mod — ₁ , and ONU ₂ , ONU ₃ is Mod — ₂ . Assume that symbol encoding is performed according to. In this case, when symbol encoding is performed in the order in which the downlink signal arrives at the data buffer 401, as shown in the time chart 43-1, the symbol encoding system is frequently switched, and the symbol encoder changeover switch 403, The number of switching at 405 increases.

  Therefore, the downlink signal scheduling unit 41 performs scheduling for the symbol encoding input rearranger 402 in order to reduce the frequency of switching the symbol encoding scheme. Symbol encoding is performed by reordering ONUs using the same symbol encoding method in the same group and performing reordering in the symbol encoding input reordering unit 402 so that data for ONUs in the same group is continuous in the time direction. By scheduling the user groups that use the same symbol encoding scheme, the switching frequency of the symbol encoding scheme during adaptive modulation is reduced as shown in the time chart 43-2.

(IFFT calculation accuracy switching frequency reduction)
In the present embodiment, the signal generation method is the calculation accuracy required for signal generation and demodulation. The slot allocation step refers to the group information of the group formed in the group formation step and the downlink use band information for the ONU 3a, minimizes the switching frequency of the calculation accuracy, and generates a signal with the required calculation accuracy Maximize the number.

  The calculation accuracy required for the IFFT calculation changes according to the transmission loss in the optical transmission line between the OLT 3b and the ONU 3a. As a method for reducing power consumption by calculation accuracy control, there is a method for reducing power consumption by reducing the number of effective bits (Patent Document 1), and IFFT calculation is performed by dynamically controlling the number of effective bits in IFFT calculation according to transmission loss. The power consumption of the device can be reduced.

  It is possible to reduce power consumption while correctly transmitting and receiving signals by generating a signal with a small number of effective bit digits for users with small transmission loss and a large number of significant bit digits for users with large transmission loss. .

  Here, when signals are transmitted to a plurality of ONUs 3a in the same time slot, and the required number of effective bit digits differs depending on the ONU 3a, the calculation accuracy required for generating the downlink signal is the required number of effective bit digits. Needs to match the largest user. Therefore, even if the ONU 3a requires a small number of effective bit digits, the number of bit digits cannot be reduced, resulting in a reduction in power consumption reduction effect.

  In order to avoid such a problem, time and carrier allocation scheduling for assigning a user having a large number of required effective bits and a user having a small effective bit number to different time slots is performed. This increases the chances of reducing the number of effective bit digits in the bit variable IFFT arithmetic unit, and improves the power consumption reduction efficiency.

  FIG. 5 is a diagram for explaining downlink signal scheduling for improving the power consumption reduction efficiency in the bit variable IFFT arithmetic unit 51 by assigning the time and carrier wave to the downlink signal. The time and carrier allocation apparatus 50 includes a data buffer 501 and a time and carrier allocator 502. The downlink signal scheduling unit 52 is based on the group information 53 and the downlink use band information 54, and the bit and variable IFFT. Scheduling information is given to the arithmetic unit 51.

Here, the data addressed to ONU ₁ , ONU _2, ONU _3, ONU ₄ sequentially arrives at the time and carrier allocation apparatus 50, and ONU ₁ , ONU ₃ receives X-bits IFFT calculation, ONU ₂ , ONU ₄ Assume that signal generation is performed by IFFT calculation of Y-bits. In this case, when signals are generated by IFFT calculation in the order of arrival of the downlink signal in the data buffer 501, time for generating an OFDM signal by IFFT calculation by Y-bits having a large number of effective bit digits as shown in the time chart 55-1 The number of slots increases.

  The downlink signal scheduling unit 52 acquires group information 53 and downlink use band information 54. As the group information 53, effective bit digit number information necessary for generating a downstream signal for each ONU is used. The downlink signal scheduling unit 52 performs scheduling so that an operation with the same number of effective bits is allocated to the same time slot, and sends scheduling information to the time and carrier allocator 502. The time and carrier allocator 502 acquires downlink traffic data from the data buffer 501 according to the given scheduling information, and uses the same calculation accuracy, that is, the time at which transmission / reception of signals to / from ONUs belonging to the same group is continued in the time direction. Also, allocation is made to the carrier wave slot, and the downlink traffic data is sent to the bit variable IFFT arithmetic unit 51. At the same time, the downlink signal scheduling unit 52 sends information related to the number of effective bit digits used for the calculation to the bit variable IFFT calculation device 51.

  When time and carrier allocation is performed in the order in which the downlink signal arrives at the time and carrier allocation apparatus 50, as shown in the time chart 55-1, data that requires different effective bit digits is included in the same time slot. there is a possibility. In such a case, it is necessary to match Y-bits having a large number of effective bit digits. In the time chart 55-1, it is possible to generate a downlink signal using the number of effective bit digits X-bits only in two time slots. On the other hand, by grouping ONUs that require the same number of significant bit digits and performing time and carrier allocation scheduling for each group, the number of significant bit digits is reduced in five time slots as shown in the time chart 55-2. It becomes possible. As a result, power consumption reduction efficiency can be improved by time and carrier allocation scheduling.

  The downlink signal scheduling unit 52 refers to the group information and rearranges the downlink data arriving at the data buffer 501 according to the group. In FIG. 5, rearrangement is performed so that the group having a smaller effective bit number is changed from a group having a smaller effective bit number. In the example of FIG. 5, the groups that require X-bits in the IFFT operation are ONU1 and ONU3, which are 12 slots in total, and the groups that require Y-bits are ONU2 and ONU4, which are 9 slots in total. In the example of FIG. 5, scheduling is performed by narrowing down in the time direction while using all available carriers.

  Even when ONUs that require the same number of significant bits are grouped, when scheduling is performed with downstream signals packed in the time direction, groups that require different number of significant bits at the same time are mixed. there is a possibility. In such a case, groups having a similar number of required effective bit digits are assigned to the same time slot. In FIG. 5, only Y-bits and X-bits are described, but it is assumed that there is an ONU that requests Z-bits by IFFT calculation. Let Z-bits> Y-bits> X-bits. It is assumed that empty slots are generated at the same time as a result of scheduling with downstream signals to the ONU requesting Z-bits being narrowed in the time direction. In this case, it is not an ONU downlink signal that requires X-bits with a small number of required effective bit digits (a signal with a large excess of calculation accuracy), but an ONU downlink signal that requires Y-bits with a similar required effective bit digit number. Scheduling is performed so that (signals with excessive calculation accuracy but close to each other) are allocated to the empty slots. By scheduling in this way, it is possible to minimize the degree of excessive calculation accuracy when a large number of effective bit digits is given to a signal having a small number of required effective bit digits.

FIG. 6 shows a case where ONUs having the same number of effective bit digits required for IFFT calculation when generating an OFDM downlink signal are grouped and scheduling is performed so as to generate a downlink signal in the same time slot for the grouped ONUs. It is the figure which showed the example of time and carrier wave allocation. Assume that the number of effective bit digits of the bit variable IFFT arithmetic unit can be reduced when generating downstream signals for ONU ₁ and ONU ₃ . When the scheduling shown in FIG. 1 is performed, it is possible to reduce the number of effective bit digits in the time slots t ₁ and t ₄ . On the other hand, as in the present embodiment, ONUs having the same effective bit digit number are grouped, and scheduling is performed so that signals are generated in the same time slot for the grouped ONUs, so that time slots t ₁ , t _{2, t 3,} can be reduced effective bit digits in _{t 9,} t _10.

Also, since scheduling is performed in the time direction, the time slot T _idle that does not require IFFT calculation is maximized.

[Uplink scheduling]
FIG. 7 is a diagram for explaining an uplink signal scheduling portion that realizes reduction in the number of switching of symbol decoding schemes and improvement in power consumption reduction efficiency by calculation accuracy control in the optical transmission system 451 described in FIG.

Uplink signal scheduling for realizing time and carrier allocation for uplink signals includes parallel-to-serial converter 70, symbol decoder 71, time and carrier deallocator 72, bit variable FFT arithmetic unit 73, data buffer 74, serial parallel. An OLT 7b having a conversion device 75, an optical signal receiving device 76, an upstream signal scheduling unit 77, request bandwidth information 78, and group information 79;
Light consisting of a plurality of ONUs (7a-1 to 7a-n) forming a group based on the calculation accuracy required in the symbol encoding method used, symbol decoding method, or FFT operation performed for OFDM signal decoding In the transmission system 451,
The uplink signal scheduling unit 77 performs uplink signal allocation control so that data using the same symbol decoding scheme is continuous in the time direction, and the uplink signals decoded by the same effective bit digit number FFT operation are the same. By assigning time and carrier to be assigned to a time slot of
The scheduling is characterized by reducing the switching frequency of the symbol decoding system and improving the effect of reducing the power consumption required for the FFT operation.

  Here, the feature of the uplink signal scheduling unit 77 is that the signal decoding method can be changed according to the communication partner such as the symbol decoding method and the number of effective bit digits used in the FFT operation required for the OFDM signal decoding. The ONU 7a having a parameter is grouped with ONUs using the same signal decoding method to reduce the switching frequency of the signal decoding method.

  In an optical transmission system having the bit variable FFT arithmetic unit 73, the power consumption reduction efficiency of the bit variable FFT arithmetic unit 73 used for OFDM signal decoding is improved by scheduling to increase the chances of reducing the number of effective bit digits. It is possible to make it.

(Reduction of symbol coding switching frequency)
In the present embodiment, the signal generation method is a symbol encoding method. In the slot assignment step, the slot is assigned to the ONU 7a by referring to the group information of the group formed in the group formation step and the uplink request bandwidth information from the ONU 7a so as to minimize the switching frequency of the symbol encoding method. Notice.

  FIG. 8 is a diagram illustrating uplink signal scheduling for reducing the frequency of switching symbol decoding schemes for uplink signals. The symbol decoding device 80 includes changeover switches (801, 803) and symbol decoders (802-1 to 802-n) that realize different symbol decoding schemes, and the uplink signal scheduling unit 81 is based on the group information 82. Scheduling information is given to the changeover switches (801, 803), and the assigned time and carrier wave information 83 are notified to the ONU 7a.

Here, the uplink data from ONU ₁ , ONU _2, ONU _3, ONU ₄ arrives sequentially at the symbol decoding device 80, ONU ₁ , ONU ₄ performs symbol decoding by Mod_1, and ONU ₂ , ONU ₃ mod_ ₂ Suppose that symbol decoding is performed according to. In this case, when symbol encoding is performed in the order in which the uplink data arrives at the symbol decoding device 80, as shown in the time chart 84-1, the symbol decoding system is frequently switched, and the symbol decoder changeover switch The switching overhead at (801, 803) increases.

  Therefore, the uplink signal scheduling unit 81 refers to the group information 82 and performs uplink signal scheduling so that the same symbol decoding scheme is continuous in the time direction. Scheduling so that the same symbol decoding scheme is continued can reduce the switching frequency of the symbol decoding scheme when adaptive modulation is performed, as shown in the time chart 84-2.

(IFFT calculation accuracy switching frequency reduction)
In the present embodiment, the signal generation method is the calculation accuracy required for signal generation and demodulation. The slot allocation step refers to the group information of the group formed in the group formation step and the uplink request bandwidth information from the ONU 7a, minimizes the switching frequency of the calculation accuracy, and performs demodulation with the required calculation accuracy Maximize the number.

  As for the upstream signal, similarly to the downstream signal, the power consumption can be reduced by reducing the number of effective bit digits of the FFT arithmetic unit used for transmission / reception of the OFDM signal. Here, when a signal is received in the same time slot from a user having a large number of effective bit digits of FFT operation and a user having a small number of effective bit digits required for decoding the signal, it is necessary to demodulate the FFT operation according to the user having a large number of effective bit digits It becomes. By allocating a user having a large number of effective bit digits and a user having a small number of valid bits to different time slots, the opportunity to reduce the number of valid bit digits in the bit variable FFT arithmetic unit increases, and the power consumption reduction efficiency is improved.

  FIG. 9 is a diagram showing uplink signal scheduling for improving the power consumption reduction efficiency in the bit variable FFT arithmetic unit 91 by time and carrier allocation for the uplink signal. The uplink signal scheduling unit 93 gives scheduling information to the time and carrier allocation deallocating device 90 and the bit variable FFT computing device 91 based on the requested bandwidth information 94 and the group information 95. Also, the allocated time and carrier wave information 96 are notified to the ONU.

It is assumed that an upstream signal that requires an X-bit FFT operation from ONU ₁ and ONU ₃ and an Y-bits FFT operation from ONU ₂ and ONU ₄ has reached bit variable FFT operation unit 91. Here, when upstream signal scheduling is performed in the order of ONU ₁ , ONU _2, ONU _3, ONU ₄ in the time direction, as shown in the time chart 97-1, Y having a large number of effective bit digits The number of time slots for decoding the OFDM signal is increased by the FFT operation using -bits.

  Therefore, uplink signal scheduling is performed so as to increase the number of time slots calculated with a small number of effective bit digits. The uplink signal scheduling unit 93 acquires the requested bandwidth information 94 and the group information 95. The group information 95 is information obtained by grouping ONUs having the same number of effective bit digits required to decode the OFDM signal transmitted from each ONU. The uplink signal scheduling unit 93 performs the scheduling so that ONUs having the same number of significant bits are assigned to the same time slot and ONUs requiring the same calculation accuracy are continuous in the time direction. The uplink signal scheduling unit 93 notifies the time and carrier information 96 to the ONU.

  Further, the uplink signal scheduling unit 93 transmits information on time and the number of significant bit digits to the bit variable FFT arithmetic unit 91. The bit variable FFT arithmetic unit 91 extracts the upstream signal from the data buffer 92 as shown in the time chart 97-2 according to the scheduling. Then, the bit variable FFT arithmetic unit 91 decodes the signal by the FFT calculation based on the number of effective bit digits necessary for decoding. This increases the time for calculation with a small number of effective bits and improves the power consumption reduction efficiency of the FFT calculator.

  Further, the scheduling information is sent to the time and carrier deallocating device 90. The time and carrier deassignment device 90 demodulates the uplink data signal assigned to the time and carrier slot, and transmits it to the host device.

  Even when ONUs that require the same number of valid bit digits are grouped, there may be a mixture of groups that require different number of valid bit digits in the same time slot. In such a case, ONU groups having a similar number of required effective bit digits are assigned to the same time slot. In FIG. 9, only Y-bits and X-bits are described. However, similarly to the description of FIG. 5, it is assumed in this embodiment that there is an ONU that requests Z-bits by IFFT calculation (Z-bits> Y-bits> X-bits). Scheduling is performed so that a downstream signal of an ONU requesting Y-bits having a near required number of valid bit digits is allocated to the empty slot, not a downstream signal of an ONU requesting X-bits having a far required number of valid bit digits. By scheduling in this way, it is possible to minimize the degree of excessive calculation accuracy when a large number of effective bit digits is given to a signal having a small number of required effective bit digits.

The following is a summary of the bandwidth allocation method and bandwidth allocation device of this embodiment.
<Issues>
The present invention reduces the frequency of switching between the symbol encoding method and the symbol decoding method, and improves the power consumption reduction efficiency by controlling the calculation accuracy of IFFT used for OFDM signal generation and FFT used for decoding And provides allocation scheduling for carrier slots.
<Solution>
In the present invention, a group is formed according to the symbol encoding method, the symbol decoding method, the calculation accuracy required for signal generation and decoding, and the time and the time so that users belonging to the same group are continuous in the time direction. Perform carrier wave allocation control.

(1): In an optical transmission system including a parent node and one or more child nodes connected to an optical fiber via a splitter,
The system assigns a slot divided into a time direction and a carrier wave direction to a child node;
Forming a plurality of groups by any method;
A time and carrier wave assignment method comprising the steps of assigning child nodes belonging to the same group to the same time slot and continuously arranging them in the time direction.

(2): The time and carrier wave allocation method of (1) above,
Grouping child nodes having the same symbol encoding scheme;
A time and carrier wave allocation technique comprising a step of arranging time and carrier wave assignments of child nodes belonging to the same group so as to be continuous in the time direction.

(3): The allocation method of (2), wherein the group information of the child node and the downlink bandwidth information for the child node are referred to;
A time and carrier wave allocation method for a downlink signal comprising a step of performing time and carrier wave allocation so that the switching frequency of the symbol coding system is minimized.

(4): Time and carrier wave allocating device for downlink signal realizing the above (3).

(5): The allocation method according to (2), wherein the group information of the child node and the uplink request bandwidth information from the child node are referred to;
A time and carrier wave allocation method for an uplink signal comprising a step of assigning a time and a carrier wave so that the switching frequency of the symbol coding system is minimized.

(6): Time and carrier wave allocating device for uplink signals that realize (5) above.

(7): The allocation method of (1) above,
Grouping child nodes with the same computational accuracy required for signal generation and demodulation;
A time and carrier wave assignment method comprising the steps of assigning child nodes belonging to the same group to the same time slot and continuously arranging them in the time direction.

(8): The allocation method of (7) above,
A step of referring to group information of child nodes and downlink use band information for the child nodes;
Minimizing the switching frequency of calculation accuracy used for signal generation;
Maximizing the number of time slots for signal generation with low computational accuracy;
A time and carrier wave allocation method for a downlink signal consisting of

(9): Time and carrier wave allocating device for the downlink signal realizing the above (8).

(10): The allocation method of (7) above,
A step of referring to the group information of the child node and the uplink request bandwidth information from the child node;
Minimizing the switching frequency of calculation accuracy used for signal decoding;
Maximizing the number of time slots for signal decoding with low computational accuracy;
A time and carrier wave allocation method for an upstream signal comprising:

(11) A time and carrier wave allocation device for an uplink signal that realizes the above (10).

<Effect>
The number of switching between the symbol encoding method and the symbol decoding method is reduced, and the band use efficiency can be improved. In addition, the number of cases where IFFT computation is performed with higher accuracy than the computation accuracy required by the signal is reduced, and the power reduction efficiency is improved.

3a, 3a-1, 3a-2, ..., 3a-n: ONU
7a, 7a-1, 7a-2, ..., 7a-n: ONU
10a, 10a-1, 10a-2, ..., 10a-n: ONU
3b, 7b, 10b: OLT
3c, 7c, 10c: Optical fibers 3d, 7d, 10d: Splitter 30: Data buffer 31: Series-parallel converter 32: Symbol encoder 33: Time and carrier allocation unit 34: Bit variable FFT processor 35: Parallel / serial conversion Apparatus 36: Optical signal transmission apparatus 37: Downlink signal scheduling section 38: Downlink use band information 39: Storage section (group information)
40: Symbol encoding device 41: Downlink signal scheduling unit 42: Storage unit (group information)
401: Data buffer 402: Symbol encoding input rearranger 403: Changeover switches 404, 404-1, 404-2, ..., 404-n: Symbol encoder 405: Changeover switch 50: Time and carrier wave allocation device 51: Bit variable IFFT arithmetic unit 52: Downlink signal scheduling unit 53: Storage unit (group information)
54: Downlink use band information 501: Data buffer 502: Time and carrier allocator 70: Parallel / serial converter 71: Symbol decoder 72: Time and carrier allocator 73: Bit variable FFT processor 74: Data buffer 75: Series-parallel converter 76: optical signal receiver 77: uplink signal scheduling unit 78: required bandwidth information 79: storage unit (group information)
80: Symbol decoding device 81: Uplink signal scheduling unit 82: Storage unit (group information)
83: Time and carrier wave information 801: Changeover switches 802, 802-1, 802-2,..., 802-n: Symbol decoder 803: Changeover switch 90: Time and carrier wave deallocation device 91: Bit variable FFT operation Device 92: Data buffer 93: Upstream signal scheduling unit 94: Request band information 95: Storage unit (group information)
96: Time and carrier information 100: Parallel / serial converter 101: Symbol code decoder 102: Time and carrier allocation / deletion device 103: Bit variable FFT operation devices 104, 104a, 104b: Data buffer 105: Series / parallel converter 106 : Optical signal transmitter / receiver 107: Signal scheduling unit 108 a: Downlink use band information 108 b: Request band information 109: Storage unit (group information)
301: Band allocation device 302: Signal generation device 451: Optical transmission system

Claims (8)

A bandwidth allocation method in which a bandwidth of an optical transmission system in which a parent node and one or more child nodes are connected by an optical transmission line is divided into slots along a time axis and an optical frequency axis, and the slots are allocated to communication with the child nodes. And
A group forming step of grouping the child nodes having the same signal generation method of signals transmitted and received between the parent node and the child notebook, and generating group information of the grouped child nodes;
When the signals of the child nodes are collectively assigned to a plurality of optical frequency slots at the same time for each group formed in the group formation step, the optical frequency at the next time when the signals do not fit in the plurality of optical frequency slots at the same time A slot assignment step to assign to the slot;
Bandwidth allocation method.   The band allocation method according to claim 1, wherein the signal generation method is a symbol coding method. In the slot allocation step,
3. The slot is allocated with reference to group information of the group formed in the group formation step and downlink use band information for the child node so that the switching frequency of the symbol coding system is minimized. Bandwidth allocation method described in 1. In the slot allocation step,
Referring to the group information of the group formed in the group formation step and uplink request bandwidth information from the child node, the slot is allocated so that the switching frequency of the symbol encoding method is minimized, and the child node is notified. The bandwidth allocation method according to claim 2, wherein:   The band allocation method according to claim 1, wherein the signal generation method has a calculation accuracy required for signal generation and demodulation. In the slot allocation step,
Referring to the group information of the group formed in the group formation step and the downlink use band information for the child node, the switching frequency of the calculation accuracy is minimized, and the number of slots for generating the signal with the required calculation accuracy is maximized. The bandwidth allocation method according to claim 5, wherein: In the slot allocation step,
Referring to the group information of the group formed in the group forming step and the uplink requested bandwidth information from the child node, the frequency of switching the calculation accuracy is minimized, and the number of slots for demodulation with the required calculation accuracy is maximized The bandwidth allocation method according to claim 5, wherein: A storage unit that executes a group formation step of the bandwidth allocation method according to any one of claims 1 to 7 and holds the group information;
A signal scheduling unit that executes a slot allocation step of the bandwidth allocation method according to any one of claims 1 to 7,
With
A bandwidth allocation device arranged in the parent node.

Images