An optical transmission device and method based on an MRDS code with tag protection
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- HUAWEI TECH CO LTD
- Filing Date
- 2025-01-30
- Publication Date
- 2026-08-06
Smart Images
Figure EP2025052329_06082026_PF_FP_ABST
Abstract
Description
[0001] AN OPTICAL TRANSMISSION DEVICE AND METHOD BASED ON AN MRDS CODE WITH TAG PROTECTION
[0002] TECHNICAL FIELD
[0003] The present disclosure is related to optical transmission, for example, transmission using optical intensity modulation and direct detection (IM / DD) based on pulse-amplitude modulation (PAM). The disclosure provides an optical transmission device, an optical transmission method, and a computer program. These may respectively be used for implementing an optical transmission scheme, which uses a minimum running digital sum (MRDS) type of code with tag protection.
[0004] BACKGROUND
[0005] In optical transmission systems, connectors may reflect a part of a transmission signal. When the reflected signal encounters a second connector along a backward propagation path, it may be reflected again. This may result in an attenuated echo, which can reach the receiver along with the main transmission signal. Said echo usually has a long delay and adds, after photodetection, a slowly varying offset onto the signal, due to a carrier-carrier beating effect. The beating generates interference in a low-frequency region, whose bandwidth depends on the linewidth of the used laser. In general, multiple reflections are also possible and would result in stronger interference.
[0006] MRDS type of codes are line codes that, when applied at the transmitter side, may deplete the signal power transmitted in the low-frequency region. Effectively, these codes mitigate multi-path interference (MPI) degradation by reducing the power transmitted in the impaired spectral region.
[0007] The MRDS codes typically operate on fixed-length data blocks (or frames), by inverting the sign of all symbols in the data block with a symbol index larger or equal than a specific value, which is called the "flipping value" (the value being also referred to as “flipping point”). The flipping value can be determined data block by data block (or frame by frame), in order to obtain a DC balanced transmission signal. The flipping value can be encoded in a short bit sequence, which is called a "tag" and is appended to the data block (or frame). The tag can be used at the receiver side to recover the original signal. After the MRDS encoding, the sum over the signal points is approximately zero when assuming symmetric transmission levels around 0 (e.g., [-3, -l, l, 3] forPAM4).
[0008] For PAM signals, the MRDS codes typically use the outer PAM symbols to encode the symbols of the tag. For systems operating at a low bit-error rate (BER), e.g., conventional Datacom systems with KP4 forward error correction (FEC), this is sufficient for protecting the tag against accidental errors. However, for other systems operating at higher BER, like optical access systems, this protection may not be enough.
[0009] Additionally, the tag adds overhead to the signal, and while MRDS codes are (almost) balanced, the tag itself may also reintroduce a DC unbalance.
[0010] SUMMARY
[0011] In view of the above, an objective of this disclosure is to provide an improved optical transmission scheme based on an MRDS type of code. An objective, for example, is to use an MRDS code without the above-mentioned issues. For instance, an objective is to overcome the issue of the accidental errors on the tag on high-BER systems. Further objectives of this disclosure includereducing the overhead introduced by tags, minimizing the unbalance that results from tag compression, and reducing DC offsets introduced by tags.
[0012] These and other objectives are achieved by the subject matter described in the independent claims. Advantageous implementations are further described in the dependent claims.
[0013] A first aspect of this disclosure provides an optical transmission device configured to: obtain one or more data blocks, each data block comprising a plurality of symbols; determine a respective flipping value for each of the data blocks; invert a sign of each symbol of each data block, which has a symbol index that is equal to or larger than the respective flipping value of that data block, to obtain a respective modified data block for each data block; determine one or more tag symbols based on the one or more determined flipping values; encode the one or more tag symbols using an error correction code (ECC); and append the one or more encoded tag symbols or parity-check symbols resulting from the encoding of the one or more tag symbols to the one or more modified data blocks, to obtain a transmission data block.
[0014] The optical transmission device of the first aspect may be used to implement an optical transmission scheme that uses a MRDS type of code in an improved manner. For instance, by adding the (e.g. short) ECC on the tag symbols (short “tag”) that encode the flipping value of the MRDS code, the flipping value is protected, reducing errors. Further, an encoding rule of encoding the flipping value into the tag symbols may be selected, to reduce or minimizes the mean value and the variance of the DC offset introduced by the tag.
[0015] In an implementation form of the first aspect, the error correction code is a single parity check code. Accordingly, a short and simple code can be used to implement the optical transmission scheme of this disclosure.
[0016] In an implementation form of the first aspect, the optical transmission device is further configured to interleave each transmission data block before optical transmission. An interleaver may be added for this purpose, and spreads the symbols of the tag along the transmission data block. This increases the robustness of the ECC to burst errors.
[0017] In an implementation form of the first aspect, the optical transmission device is configured to add the parity-check symbols resulting from a joint encoding of multiple tag symbols using the ECC to the one or more modified data blocks.
[0018] In an implementation form of the first aspect, the optical transmission device is configured to determine the one or more tag symbols based on cost-encoding the one or more flipping values. Cost-encoding may mean that the optical transmission device selects the mapping between tag symbols and flipping values by evaluating a “cost function”. The cost function in this case may be tied to maintaining signal balance (e.g., DC balance) or minimizing the probability of transmission errors. By assigning the tag symbols with the lowest “cost” to the most likely flipping values, the optical transmission device may optimize the overall signal integrity, and may reduce errors during high-speed optical transmission.
[0019] In an implementation form of the first aspect, the optical transmission device is configured to, in order to determine the respective flipping value for a data block: calculate a plurality of flipping value candidates; and determine the flipping value based on the flipping value candidates using a cost function.
[0020] In an implementation form of the first aspect, the optical transmission device is further configured to compress the one or more determined flipping values before encoding them with the ECC and before inverting the sign of each symbol of each data block.In an implementation form of the first aspect, the optical transmission device is configured to compress the one or more determined flipping value by lossy compression.
[0021] In an implementation form of the first aspect, the optical transmission device is configured to compress the one or more determined flipping values by non-uniform quantization.
[0022] A quantization in the choice of the flipping values may be added to reduce the overhead of the MRDS or similar code. Non-uniform quantization is a way of converting continuous or high-resolution digital values into fewer discrete values where the spacing between these values varies instead of being uniform. This may give finer resolution (more values) for ranges where precision matters most, enabling more efficient compression and better quality with fewer symbols.
[0023] In an implementation form of the first aspect, the optical transmission device is further configured to determine one or more next flipping values for one or more next data blocks based on the transmission data block.
[0024] In an implementation form of the first aspect, the optical transmission device is configured to determine the one or more next flipping values based on a minimized cost function applied to the transmission data block and the one or more next data blocks. A typical choice of the cost function is DC unbalance. In this case, the goal is to compensate the DC unbalance of the current transmission block with the unbalance of the next transmission block. Generally, the cost function applied to current and next transmission blocks, or a sequence of transmission blocks, is minimized.
[0025] DC unbalance refers to a condition, in which the average value (or “direct current” (DC) component) of a transmission signal is not zero. This may happen when more positive symbols than negative symbols (or vice versa) are transmitted over time, creating a DC offset.
[0026] In an implementation form of the first aspect, the optical transmission device is further configured to: jointly encode multiple sets of tag symbols determined for multiple data blocks; and add the jointly encoded tag symbols to the transmission data block, optionally after interleaving.
[0027] In an implementation form of the first aspect, the symbols are 4-level pulse amplitude modulation (PAM4) symbols.
[0028] A second aspect of this disclosure provides an optical transmission method comprising: obtaining one or more data blocks, each data block comprising a plurality of symbols; determining a flipping value for each of the one or more data blocks; inverting a sign of each symbol of each data block, which has a symbol index that is equal to or larger than the respective flipping value for that data block, to obtain a respective modified data block per data block; determining one or more tag symbols based on the one or more determined flipping values; encoding the one or more determined tag symbols using an ECC; and appending the one or more encoded tag symbols to the one or more modified data blocks, to obtain a transmission data block.
[0029] The optical transmission method of the second aspect can have implementation forms that correspond to the implementation forms of the optical transmission device. The optical transmission method may have the same advantages as the optical transmission device.
[0030] A third aspect of this disclosure provides a computer program comprising instructions which, when the program is executed by a processor of an optical transmission device, cause the optical transmission device to perform the method according to the first aspect or any implementation forms thereof.A fourth aspect of this disclosure provides a non-transitory storage medium storing executable program code which, when executed by a processor, causes the method according to the second or any of its implementation forms to be performed.
[0031] In summary, this disclosure proposes a new method based on an MRDS type of code - a code, which inverts signs of symbols of a data block with indices larger than a flipping value - which may ensure reliable transmission also in case of poor channel quality. The tag of the code that encodes the flipping value is protected with a short ECC. To increase the robustness of the ECC to burst errors, its symbols (e.g. bits) may be spread along the transmission data block by using an interleaver.
[0032] Furthermore, to reduce the overhead of the code, quantization of the flipping value may be applied. Since, after quantization, each resulting flipping value may be suboptimal, a sequence of flipping values may be chosen that, in average, cancels the quantization error. Additionally, an encoding rule may be used for encoding the flipping value into tag symbols, which reduces or minimizes the average and the variance of the unbalance of the tag itself.
[0033] It has to be noted that all entities, elements, units and means described in the present application could be implemented by software or hardware elements or any kind of combination thereof. All steps performed by the various entities described in the present application, as well as the functionalities described to be performed by the various entities, are intended to mean that the respective entity is adapted to or configured to perform the respective steps and functionalities.
[0034] Even if, in the following description of specific embodiments, a specific functionality or step to be performed by external entities is not reflected in the description of a specific detailed element of that entity, which performs that specific step or functionality, it should be clear for a skilled person that these methods and functionalities can be implemented by respective software or hardware elements, or any kind of combination thereof.
[0035] BRIEF DESCRIPTION OF DRAWINGS
[0036] The above described aspects and implementation forms are explained in the following description in relation to the enclosed drawings, in which:
[0037] FIG. 1 shows an optical transmission device according to this disclosure.
[0038] FIG. 2 shows a first exemplary embodiment of an optical transmission device according to this disclosure.
[0039] FIG. 3 shows an interaction example of a single parity check code (as an example of the ECC) in the tag, and an interleaver on a symbol sequence of a data block.
[0040] FIG. 4 shows exemplarily a probability mass function (PMF) of a flipping index that depends on the MRDS algorithm and the choice of the flipping value.
[0041] FIG. 5 shows a table with exemplary flipping values (indices), probabilities, and labels.
[0042] FIG. 6 shows a second exemplary embodiment of an optical transmission device according to this disclosure.
[0043] FIG. 7 shows a third exemplary embodiment of an optical transmission device according to this disclosure.
[0044] FIG. 8 shows an optical transmission method according to this disclosure.DETAILED DESCRIPTION OF EMBODIMENTS
[0045] Illustrative but exemplary embodiments of the optical transmission device and method are described in the following with reference to the above-mentioned figures. Although the description provides also details of the embodiments, it should be noted that even these embodiments are intended to be exemplary, and in no way limiting to the scope of the disclosure.
[0046] FIG. 1 shows an optical transmission device 100 according to this disclosure. The optical transmission device 100 may be a transceiver, for instance, an intensity modulation and direct detection (IM / DD) based transceiver. The optical transmission device 100 may be a device in a fiber-based optical communication system, e.g., in a passive optical network (PON) or in a datacenter network (DCN). The optical transmission device 100 could be an optical line terminal (OLT) or an optical network unit (ONU) of a PON, or a network device of a DCN, or the like. Generally, the optical transmission device 100 can be any device that is capable of optical transmission.
[0047] The optical transmission device 100 is configured to obtain one or more data blocks 101, wherein each data block 101 comprises a plurality of symbols. The symbols could be bits, but may also be PAM symbols. For instance, the optical transmission device 100 may receive the one or more data blocks 101, for example, from a symbol source or from another device. The optical transmission device 100 could also by itself generate the one or more data blocks 101.
[0048] The optical transmission device 100 is further configured to determine (at block 111 ) a respective flipping value 102 for each of the data blocks 101, i.e., a flipping value 102 per obtained data block 101. The flipping value 102 denotes the flipping index, and thus the flipping point, similarly as in a conventional optical transmission scheme based on an MRDS code. The optical transmission device 100 may also use such an MRDS code. The respective flipping value 102 could be obtained by first calculating a plurality of flipping value candidates, and then determine the flipping value 102 based on the flipping value candidates using, for example, a cost function.
[0049] The optical transmission device 100 is further configured to invert (at block 112) a sign of each symbol of each data block 101, which has a symbol index that is equal to or larger than the respective flipping value 102 of that data block 101. This may be done according to the MRDS code. A respective modified data block 103 is thereby obtained for each data block 101, i.e., a modified data block 103 per data block 101.
[0050] The optical transmission device 100 is further configured to determine (at block 113) one or more tag symbols 104 (“tag”) based on the one or more previously determined flipping values 102. The tag symbols 104 may encode the flipping value 102. The tag symbols 104 can be determined as for a conventional scheme using MRDS code, or could be determined based on a cost-encoding of the one or more flipping values 102.
[0051] The optical transmission device 100 is further configured to encode (at block 114) the one or more tag symbols 104 using an error correction code (ECC). The ECC can be a parity check code, specifically, a single parity check code may be used. This ECC may be used to protect the tag symbols 104, and thus the encoded flipping value(s) 102.
[0052] The optical transmission device 100 is further configured to append (at block 115) the one or more encoded tag symbols 105, or to append parity-check symbols 105 resulting from the encoding (at block 114) of the one or more tag symbols 104, to the one or more modified data blocks 103. Thereby - or optionally after further processing steps explained later, like interleaving - a transmission data block 106 may be finally obtained, which can be transmitted optically by an optical transmitter of the optical transmission device.The optical transmission device 100 may comprise a processor (not shown) or processing circuitry to perform, conduct or initiate various operations described above. For instance, each of the blocks 111 - 115 could be implemented by suitable processing circuitry. That is, each block 111 - 115 may be processing circuitry. The functions of all the blocks 111 - 115 could be implemented by a single processor or controller of the optical transmission device 100, for example, a processor that comprises dedicated processing circuitry for each functional block 111 - 115.
[0053] Such processing circuitry may comprise hardware and / or may be controlled by software. The hardware may comprise analog circuitry or digital circuitry, or both analog and digital circuitry. The digital circuitry may comprise components such as application-specific integrated circuits (ASICs), field-programmable arrays (FPGAs), digital signal processors (DSPs), or multi-purpose processors.
[0054] The optical transmission device 100 may further comprise a memory or memory circuitry, which stores one or more instruction(s) that can be executed by the processor or by the processing circuitry, in particular under control of the software. For instance, the memory circuitry may comprise a non-transitory storage medium storing executable software code which, when executed by the processor or the processing circuitry, causes the various operations described above to be performed. In one embodiment, the processing circuitry comprises one or more processors and a non-transitory memory connected to the one or more processors. The non-transitory memory may carry executable program code which, when executed by the one or more processors, causes the optical transmission device 100 to perform, conduct or initiate the operations or methods described herein.
[0055] A solution of this present disclosure, as generally explained with reference to FIG. 1, suggests adding an ECC, e.g. a single parity-check (SPC) code, to protect the tag symbols 104 that encode the flipping value(s) 102. While generally such SPC codes are rather weak codes, surprisingly, in the present case they are enough to boost the performance by multiple decades in the BER.
[0056] FIG. 2 shows a first exemplary embodiment of an optical transmission device 100 according to this disclosure.
[0057] As a starting point for this first exemplary embodiment, a conventional 4-level pulse amplitude modulation (PAM4) transmission system may be considered. Such a transmission system operates, before forward error correction (FEC), at a biterror rate of approximately 2.0-10'2, and uses an MRDS code with data blocks (or frames) of length 256. The MRDS tag may consist of 8 bits, which are mapped to 8 PAM4 symbols, for which only the two outermost PAM4 points are used. If in this kind of transmission system, transmission impairments are modeled as additive white Gaussian noise (AWGN), the resulting BER on those outer points can be determined to be 3.4-1 O'9. Assuming that the transmission system works at 50 Gbaud, this would mean that every 101 seconds an error appears on the most significant bit of the tag. Every such error results in a long error burst, which very likely leads to residual errors after FEC decoding. Obviously, this error rate after FEC decoding is not acceptable. Therefore, from the perspective of the system, the presence of the MRDS code could make transmission unfeasible.
[0058] However, in the optical transmission device 100 according to the first exemplary embodiment, an extra symbol of redundancy is added via the single parity-check code (corresponding to block 114, and the ECC in FIG. 1). Further, maximum-likelihood decoding may be performed. In this case, the bit error rate on the tag symbols 104 drops from 3.4-10'9to 5.0-10'16. This means that an error burst of 128 bits appears approximately once every 20 years. This error rate is acceptable for virtually any practical system. Maximum-likelihood decoding of SPC codes is trivial, and can be implemented by simply flipping the least reliable bit when the parity check is not fulfilled.The optical transmission device 100 of FIG. 2 includes similar blocks 111 - 115 as described above with respect to FIG. 1. Specifically, in the first exemplary embodiment, the one or more tag symbols 104 are determined based on cost-encoding (block 211, relates to block 113) the one or more flipping values 102. Block 115 is implemented by a multiplexer (MUX). The ECC code at block 114 is implemented as SPC code.
[0059] Determining the flipping values 102 (at block 111) includes compressing the one or more determined flipping values 102. That is, the one or more determined flipping values 102 are compressed before encoding them with the ECC, and before inverting the sign of each symbol of each data block 101. The compression may be done by lossy compression and / or by non-uniform quantization, as will be explained below.
[0060] Optionally, an interleaver 212 may further be introduced (after the MUX), in order to separate the ECC / SPC symbols over time. This makes the ECC / SPC code more robust to burst errors that may be present over the transmission link, before MRDS decoding. The interleaver 212 interleaves each transmission data block 106 before optical transmission.
[0061] FIG. 3 shows an example for the interaction of an SPC code in the tag 104 and an interleaver 212 on a bit sequence of a data block 101. Specifically, the example in FIG. 3 shows how the MRDS code (inverting symbols according to the flipping value 102), the SPC code, and the interleaver work together in the case of PAM4. The original frame has a DC unbalance of 6 (=-l-3+1+1+3+3+3-1 ). The flipping value 102 is determined to be 6 (assuming that counting starts at 0). After MRDS encoding, i.e. after sign flipping, the DC unbalance is reduced to 2. The flipping value 102 is encoded, for the moment using natural binary encoding, as " 110", and is mapped to the tag (3, 3, -3). The tag 104 is extended by a single parity check to (3, 3, -3, -3), and is appended to the MRDS frame, i.e., the modified data block 103. Finally, the tag symbols 104 are spread (or distributed) over the modified data block 103.
[0062] The transmission data blocks 106 are exactly balanced. In fact, even if one optimizes the choice of the flipping point 102, alreadv the flipped sequence, i.e. the modified data block 103, is not always exactly balanced and, in general, also the tag 104 contributes to the overall DC unbalance. Therefore, at the cost of almost no additional penalty, a lossy compression can be used on the flipping value 102. When the choice of potential flipping value 102 decreases, the effect is twofold. In the one hand, the search for the flipping value 102 becomes easier, and, on the other hand, the tag 104 can be shortened, because excluded flipping values 102 do not need to be encoded.
[0063] As an example, an MRDS block of length 384 can be considered. In the absence of compression, one needs [log2384] = 9 bits to address each possible flipping value 102. It could be agreed to use as flipping values 102 only indices that, divided by 3, leave a remainder of 1, i.e., 1,4,7, ...,382. In that case, there would be only 128 possible values, which could be encoded in 7 bits. Also, during MRDS encoding, in the search for the best flipping value 102, only 128 values could be considered instead of 384. The present disclosure calls this type of compression a "coarse quantization" of the flipping value 102.
[0064] A second compression method could be to use every second value in the range of 64 to 318 (including 318), which amounts to a selection of 128 values. The present disclosure calls this approach a "coarse quantization with saturation".
[0065] A third compression method could be to use every second value in the range 128 to 255 and in the remaining region only every fourth value. This would amount again to 128 values in total. This is, like the previous example, a case of non-uniform quantization.Especially after compression of the flipping value 102, each individual MRDS sequence, i.e., each individual modified data block 103, may not be perfectly balanced. The device can remember the unbalance of the current block (e.g., the resulting transmission data block 106), and can use it at the next block to bias the new unbalance into the opposite direction.
[0066] For example, assuming that the running digital sum over the previous block and all other previous block after flipping was fft-i, and the sum over the current block before flipping is cr£, the following algorithm can be defined.
[0067] G [— |tr£|, | cr£|], select the new flipping value 102 such that the running sum is closest to (cr£+
[0068]
[0069] This guarantees that at least one crossing point exists. Otherwise, select the last or first symbol as flipping value 102 depending on which one is better. The total running sum may be updated with the new value after flipping fft.
[0070] In other words, one or more next flipping values 102 for one or more next data blocks 101 can be determined based on the transmission data block 106.
[0071] Additionally, a forgetting factor may be introduced, which scales the running digital sum over all previous blocks with a factor (1-f) with 0<f<l . Setting f to 1 would mean that the previous blocks are not considered, and setting f to 0 would mean that the exact running sum over all blocks is considered.
[0072] Further, there are in general multiple possible flipping values 102 leading to the same unbalance. MRDS encoding relies typically on some pre-determined rule to solve "ties" . For example, the encoding algorithm can choose the first possible flipping value 102, or the last one, or the one that is closest to the middle of the frame.
[0073] The possible (quantized) flipping values 102 have different probabilities. The MRDS algorithm and the choice of which flipping value to take, influence the probability mass function (PMF) of the flipping index. This distribution may be pointy, i.e., some values may appear more often than others. The optical transmission device 100 can take advantage of the shape of this PMF, so as to minimize the effect of the unbalance introduced by the MRDS tag. To this end, more balanced binary labels can be assigned to indices with a higher probability.
[0074] FIG. 4 shows exemplarily such a (PMF) of a flipping value. The PMF in FIG. 4 was obtained for PAM4 after quantization of 256 positions to 16 different indices.
[0075] The bit mapping of the flipping values 102 can be optimized in the following way:
[0076] • Order the flipping indices i according to their probability of appearance in descending order.
[0077] • Order all the labels t = T0Tt...
[0078]
[0079] according to their absolute sum S =
[0080]
[0081] | in ascending order. • Combine the j'-th element of each sorted list, to obtain the optimized bit mapping.
[0082] In this example, with reference to FIG. 5, the optimized bit mapping reduces the expected quadratic distance from the 0-sum from 25.56 (obtained with natural binary mapping) to 15.84.
[0083] The first exemplary embodiment of the optical transmission device 100 provides several advantages. For example, the error probability on the tag symbols 104 (e.g., tag bits) is reduced by using an SPC code, and optionally the interleaver 212. Further, the tag length and the encoding complexity may be reduced via quantization. Further, the overall unbalance may be reduced via cost encoding of the tag symbols 104. Further, the average unbalance over time may be reduced via flipping value selection with memory.In the following two exemplary embodiments, it is explained how to use an error correcting code to protect jointly the tag symbols 104 of multiple data blocks 101.
[0084] FIG. 6 shows a second exemplary embodiment of an optical transmission device 100 according to this disclosure. The second exemplary embodiment is based on product-like encoding.
[0085] In this second exemplary embodiment, the structure of the first exemplary embodiment is extended (the same blocks 111 - 115 as in the second exemplary embodiment are used, but multiple of them), wherein a short ECC, e.g. an SPC code, is applied individually to the tag symbols 104 of each data block 101. For example, if k is the length of each tag, the SPC encoder may extend each tag 104 to a length of / r+1, by appending a single parity-check. Additionally, a second code spans VMRDS blocks and encodes jointly their N tags. The second code (at block 611), which can also be an SPC code, may be applied k times. The first instance of the second code protects jointly the first symbols (e.g. first bits) of the V MRDS tags, the second instance protects jointly the second symbols (e.g. second bits) of the N tags, and the j-th instance (l<j < k) protects jointly the j-thbits of the N tags. The k instances of the second SPC code generate together k additional parity checks 605, which may be appended to the jV-th MRDS frame. The optical transmission device 100 can add the parity-check symbols resulting from a joint encoding 114 of multiple tag symbols 104 using the ECC to the one or more modified data blocks 103.
[0086] FIG. 7 shows a third exemplary embodiment of an optical transmission device 100 according to this disclosure. The second exemplary embodiment is based on “FEC over all words”. In this third exemplary embodiment, the structure of the first exemplary embodiment is reused, except for the ECC. Now the ECC (represented by the block 114 “Parity calculation”) is fed from the tags 104 of multiple data blocks 101 (or frames). Its parity checks are appended to the last data block. Accordingly, the optical transmission device 100 can jointly encode (at block 114) multiple sets of tag symbols 104 determined for multiple data blocks 101, and can add the jointly encoded tag symbols 105 to the transmission data block 106, optionally after interleaving.
[0087] Advantageously, for the second and third exemplary embodiments, the performance can be improved, if a more complex ECC is used in place of the SPC code.
[0088] The main application scenario of all embodiments is in a passive optical network (PON) or a datacenter network (DCNs) employing a PAM format, e.g., PAM2, PAM3, PAM4, PAM6 or PAM8.
[0089] In sum, adding a short error-correction code on the tag 104 that encodes the flipping value 102 of an MRDS code reduces the error probability, and thus allows operating the system at an increased BER. This can be improved by using further the optional interleaver 212 to spread the symbols of the tag 104 along the transmission data block 106. Other aspects of the disclosure are a quantization in the choice of the flipping values 102 to reduce the overhead of the MRDS code, and the use of an encoding rule of the flipping values 102 into bits, which reduces or minimizes the mean value and the variance of the DC offset introduced by the tag 104.
[0090] The present disclosure has been described in conjunction with various embodiments as examples as well as implementations. However, other variations can be understood and effected by those persons skilled in the art and practicing the claimed matter, from the studies of the drawings, this disclosure and the independent claims. In the claims as well as in the description the word “comprising” does not exclude other elements or steps and the indefinite article “a” or “an” does not exclude a plurality. A single element or other unit may fulfill the functions of several entities or items recited in the claims. The mere fact that certain measures are recited in the mutual different dependent claims does not indicate that a combination of these measures cannot be used in an advantageous implementation.
Claims
CLAIMS1. An optical transmission device (100) configured to:obtain one or more data blocks (101), each data block (101) comprising a plurality of symbols;determine (111) a respective flipping value (102) for each of the data blocks (101);invert (112) a sign of each symbol of each data block (101), which has a symbol index that is equal to or larger than the respective flipping value (102) of that data block (101), to obtain a respective modified data block (103) for each data block (101);determine (113) one or more tag symbols (104) based on the one or more determined flipping values (102); encode (114) the one or more tag symbols (104) using an error correction code, ECC; andappend (115) the one or more encoded tag symbols (105) or parity-check symbols resulting from the encoding (114) of the one or more tag symbols (104) to the one or more modified data blocks (103), to obtain a transmission data block (106).
2. The optical transmission device (100) according to claim 1, wherein the error correction code is a single parity check code.
3. The optical transmission device (100) according to claim 1 or 2, further configured to interleave (212) each transmission data block (106) before optical transmission.
4. The optical transmission device (100) according to one of the claims 1 to 3, configured to add the parity-check symbols resulting from a joint encoding (114) of multiple tag symbols (104) using the ECC to the one or more modified data blocks (103).
5. The optical transmission device (100) according to one of the claims 1 to 4, configured to determine (113) the one or more tag symbols (104) based on cost-encoding (211) the one or more flipping values (102).
6. The optical transmission device (100) according to one of the claims 1 to 5, configured to, in order to determine the respective flipping value (102) for a data block (101):calculate a plurality of flipping value candidates; anddetermine the flipping value (102) based on the flipping value candidates using a cost function.
7. The optical transmission device (100) according to one of the claims 1 to 6, further configured to compress (111) the one or more determined flipping values (102) before encoding (114) them with the ECC and before inverting (112) the sign of each symbol of each data block (101).
8. The optical transmission device (100) according to claim 7, configured to compress the one or more determined flipping values (102) by lossy compression.
9. The optical transmission device (100) according to claim 7, configured to compress the one or more determined flipping values (102) by non-uniform quantization.
10. The optical transmission device (100) according to one of the claims 1 to 9, further configured to determine one or more next flipping values (102) for one or more next data blocks (101) based on the transmission data block (106).
11. The optical transmission device (100) according to claim 10, configured to determine the one or more next flipping values (102) based on a minimized cost function applied to the transmission data block (106) and the one or more next data blocks (101).
12. The optical transmission device (100) according to one of the claims 1 to 11, further configured to:jointly encode (114) multiple sets of tag symbols (104) determined for multiple data blocks (101); andadd the jointly encoded tag symbols (105) to the transmission data block (106), optionally after interleaving.
13. The optical transmission device (100) according to one of the claims 1 to 12, wherein the symbols are 4-level pulse amplitude modulation, PAM4, symbols.
14. An optical transmission method (800) comprising:obtaining (801) one or more data blocks (101), each data block (101) comprising a plurality of symbols; determining (802) a flipping value (102) for each of the one or more data blocks (101);inverting (803) a sign of each symbol of each data block (101), which has a symbol index that is equal to or larger than the respective flipping value (102) forthat data block (101), to obtain a respective modified data block (103) per data block (101);determining (804) one or more tag symbols based on the one or more determined flipping values (102); encoding (805) the one or more determined tag symbols (104) using an error correction code, ECC; and appending (806) the one or more encoded tag symbols (105) to the one or more modified data blocks (103), to obtain a transmission data block (106).
15. A computer program comprising instructions which, when the program is executed by a processor of an optical transmission device (100), cause the optical transmission device (100) to perform the method (800) according to claim 14.