# EFFICIENT DATA ENCODING

Circuits, methods, and apparatus for efficiently implementing encoding and decoding between binary and multilevel data.

## Latest Apple Patents:

- Concurrent channel state information (CSI) capability reporting using multiple codebooks
- Flexible electronic subscriber identity module deployment
- SMS enhancement for new radio
- Access control for wireless cellular communication systems
- Techniques for gap-based feedback measurement in new radio wireless cellular network

**Description**

**CROSS-REFERENCES TO RELATED APPLICATIONS**

This application is a continuation of U.S. patent application Ser. No. 17/073,861, filed Oct. 19, 2020, which is a continuation of U.S. patent application Ser. No. 16/459,402, filed Jul. 1, 2019, which is a nonprovisional of, and claims the benefit of, U.S. provisional patent application No. 62/692,295, filed Jun. 29, 2018, which are incorporated by reference.

**BACKGROUND**

Computers and computing devices, such as laptops, all-in-one computers, smartphones, tablets, and other devices, perform data operations using binary data. Binary data is made up of individual bits, which can have one of two states, a 0 or a low, and a 1 or a high. These states can also be referred to as the ON and OFF states.

These computers and computing devices can communicate with each other over cables that can include a number of conductors. These conductors can convey signals, power supplies, or other voltages between or among the devices. These conductors can include wires, coaxial cables, fiber-optic cables, or other types of conductors.

Binary data can be transmitted and received using these conductors. Other types of data having more than one state—multilevel data—can also be transmitted and received using these conductors. Transmitting multilevel data as compared to binary data can increase data rates and more efficiently utilize available bandwidth. But the change from binary data to multilevel data requires an encoding circuit, while the change back from multilevel data to binary data requires a decoding circuit. These circuits can be complex and difficult to implement. These complex circuits can delay computing device deployment, that is, new product time-to-market. They can also increase power dissipation and other resource utilization.

Thus, what is needed are circuits, methods, and apparatus for efficiently implementing encoding and decoding between binary and multilevel data.

**SUMMARY**

Accordingly, embodiments of the present invention can provide circuits, methods, and apparatus for efficiently implementing encoding and decoding between binary and multilevel data.

An illustrative embodiment of the present invention can provide a method of encoding data where a first number of bits are received and encoded into a second number of symbols. The received bits can be organized into groups of bits, where one of the groups of bits can be referred to as first branch bits and another group of bits can be referred to as second branch bits. One or more first branch bits can be read, and it can be determined if the first branch bits have a value in a set of one or more possible values for the branch bits. If they do, the remaining bits can be encoded in a series of symbols. For example, groups of bits can be encoded into symbols using a translation table (or other method.) The first branch bits can also be encoded into a symbol, for example if the first branch bits include two or more bits. In these and other embodiments of the present invention, the first branch bits might not need to be encoded, for example where the first branch bits only include one bit. If the first branch bits do not have a value in the set of one or more possible values, then one or more second branch bits can be read. Given the value of the second branch bits, one or more symbols in the second number of symbols can be assigned an address value, where the address value is a value that binary data can't be encoded into using the translation table. The remaining bits can then be encoded into the remaining symbols in the second number of symbols using translation table.

Another illustrative embodiment of the present invention can provide a method of decoding data where the second number of symbols are received and decoded into the first number of bits. The received second number of symbols can be examined to see if one or more symbols have been assigned the address value. If they do, then depending on which symbols have the address value, the remaining symbols in the second number of symbols can be decoded, resulting in the decoded first number of bits. If the received second number of symbols do not include the address value, then the symbols can be decoded and the first number of bits recovered.

In a specific embodiment, eleven bits can be encoded into a series of 7 three-level or ternary symbols. For example, binary bits can be encoded into symbols that can be pulse-amplitude modulated symbols or PAM3 symbols. The first two of these bits can be used as the first branch bits. When the first branch bits have one of three possible values, for example 00, 01, or 10, the first branch bits can be encoded as a first symbol. The three subsequent groups of three bits each can then be encoded into two symbols. The resulting 7 symbols can then be transmitted. If the first branch bits do not have one of three possible values, for example, they have a value of 11, the first branch bits can't be encoded in the first symbol, since the 00, 01, and 10 states already correspond to the three possible symbol states. Instead, the first group of three bits, which can be referred to as the second branch bits, are read. Based on the value of the second branch bits, a first symbol can be assigned a value and one of the three subsequent groups of two symbols can be assigned the address value. The first branch bits do not need to be encoded since their value (11) can be determined by a decoder from the presence of an address value. The remaining bits can be encoded into the remaining symbols.

In these and other embodiments of the present invention, groups of three bits can be encoded into two three-level symbols. Three bits can have one of 8 values (8 different combinations), while two three-level symbols can have 9 possible values. This leaves a ninth state for the two symbols that is unused in the encoding of the three bits. This unused ninth state can then be used as an address value, where the presence and position of the address value can be used in decoding a series of symbols.

For example, in a specific embodiment of the present invention, 7 three-level symbols can be decoded into eleven bits. Again, these symbols can be pulse-amplitude modulated symbols or PAM3 symbols. A group of seven symbols can be received. The symbols can be checked for the presence of an address value. If an address value is found, a value for the two first branch bits can be determined to have a value of 11, consistent with the encoding scheme above. The first symbol can be read and from that, along with the location of the address bits, the second branch bits can be determined. The remaining bits can be determined given the position of the address bits. More specifically, since the first branch bits and the second branch bits are known, the 6 bits in the last two groups of three bits remain to be determined. The address value can be located in a pair of the remaining 6 symbols, leaving two groups of two symbols to be decoded. Each group of two symbols can then be decoded into three bits resulting in the last 6 bits, thereby completing the decoded set of eleven bits. If an address value is not present in the 7 three-level symbols, the first symbol can be decoded into the two first branch bits. The remaining three groups of two bits each can be decoded into three groups of three bits, thereby completing the eleven decoded bits.

Another illustrative embodiment of the present invention can provide a method of encoding data where a first number of bits are received and encoded into a second number of symbols. One or more first branch bits can be read, and it can be determined if the first branch bits have a value in a first set of one or more possible values for the branch bits. If they do, the remaining bits can be encoded in a series of symbols using a translation table (or other method.) The first branch bits can also be encoded into a symbol, for example if the first branch bits include two or more bits. In these and other embodiments of the present invention, the first branch bits might not need to be encoded, for example where the first branch bits only include one bit. If the first branch bits do not have a value in the first set of one or more possible values, then one or more second branch bits can be read. It can then be determined if the second branch bits have a value in a second set of values. If they do, then given the value of the second branch bits, one or more symbols in the second number of symbols can be assigned an address value, where the address value is a value that binary data can't be encoded into using the translation table. If the second branch bits do not have a value in the second set of values, then one or more third branch bits can be read. As before, it can then be determined if the third branch bits have a value in a third set of values. If they do, then given the value of the third branch bits, two or more symbols in the second number of symbols can be assigned an address value. This can be repeated as necessary until the remaining bits can be encoded into the remaining symbols in the second number of symbols.

Another illustrative embodiment of the present invention can provide a method of decoding data where the second number of symbols are received and decoded into the first number of bits. The received second number of symbols can be examined to see if one or more symbols have been assigned the address value. If they have, then depending on which symbols have the address value, the remaining symbols in the second number of symbols can be decoded, resulting in the decoded first number of bits. If the received second number of symbols do not include the address value, then the symbols can be decoded and the first number of bits recovered.

In a specific embodiment, 19 bits can be encoded into a series of 12 three-level symbols. As before, these binary bits can be encoded into symbols that can be pulse-amplitude modulated symbols or PAM3 symbols. The first of these bits can be used as the first branch bit. If the first branch bit has a specific value, for example 0, then the remaining 18 bits can be arranged in 6 groups of three bits each and encoded into a corresponding 6 pairs of two symbols for a total of 12 symbols. If the first branch bit does not have the specific value, for example it is a 1, then a subsequent set of three bits can be used as second branch bits. If the second branch bits have a value in a second set of values, then one pair of symbols can be assigned the address value, where the location of the address value is dependent on the value of the second branch bits. The remaining 5 groups of three bits can be encoded into the remaining 5 pairs of symbols. There is no need to encode the first branch bit or the second branch bits since their value can be determined by a decoder from the presence of an address value. If the second branch bits do not have a value in the second set of values, then a subsequent set of three bits can be used as third branch bits. If the second branch bits have a first specific value, and the third branch bits have a value in a third set of values, then two pairs of symbols can be assigned the address value, where the location of the address values are dependent on the value of the second branch bits and the third branch bits. The remaining 4 groups of three bits can be encoded into the remaining 4 pairs of symbols. Again, there is no need to encode the first branch bit since its value can be determined by a decoder from the presence of an address value. There is also no need to encode the second or third branch bits since their value can be determined by a decoder from the presence of two address values. If the second branch bits do not have the first specific value and the third branch bits have a value in a third set of values, then two pairs of symbols can be assigned the address value, where the location of the address values are dependent on the value of the second branch bits and the third branch bits. The remaining 4 groups of three bits can be encoded into the remaining 4 pairs of symbols. Again, there is no need to encode the first branch bit since its value can be determined by a decoder from the presence of an address value. There is also no need to encode the second or third branch bits since their value can be determined by a decoder from the presence of two address values. If the third branch bits do not have a value in the third set of values, then a subsequent set of three bits can be used as fourth branch bits. Three pairs of symbols can be assigned the address value, where the location of the address values are dependent on the value of the second branch bits, the third branch bits, and the fourth branch bits. The remaining 3 groups of three bits can be encoded into the remaining 3 pairs of symbols. Again, there is no need to encode the first branch bit since its value can be determined by the presence of an address value. There is also no need to encode the second, third, or fourth branch bits since their value can be determined by a decoder from the presence of three address values.

In a specific embodiment, a series of 12 three-level symbols can be decoded into 19 bits. In this example, the 12 symbols can be arranged in 6 pairs. Three of those 6 pairs can have the address value in one of 8 combinations. The 6 symbols pairs are checked to see if they have 3 pairs with address values in one of those 8 combinations. If they do, a first branch bit, second branch bits, third branch bits, and fourth branch bits can be determined. The remaining bits can be decoded from the remaining symbols. If three symbol pair do not have the address value, a check of whether two symbol pairs have the address value is done. The 6 symbol pairs can have two with address values in one of 15 combinations. If they do, a first branch bit, second branch bits, and third branch bits can be determined. The remaining bits can be decoded from the remaining symbols. If two symbol pairs do not have the address value, a check of whether one symbol pair has the address value is done. If it does, a first branch bit and second branch bits can be determined. The remaining bits can be decoded from the remaining symbols. If no symbol pair have the address value, the first branch bit can be set and the symbol pairs can be decoded directly. Again, each of these symbol decoding can be done using a translation table or other appropriate method.

These and other embodiments of the present invention can provide encoders and decoders that can be readily implemented using a minimal amount of logic gates. This can reduce component size, save power, speed design, and improve yields. While examples are shown utilizing specific numbers of bits, symbols, and type of symbols, these and other embodiments of the present invention can utilize different numbers of bits, symbols, and different types of symbol, such as four or five level symbols. Other types of encoding (and decoding), such as phase or frequency encoding, can also be used.

Embodiments of the present invention can provide data encoders and decoders that can be used in various types of devices, such as lighting equipment, portable computing devices, tablet computers, desktop computers, laptops, all-in-one computers, wearable computing devices, cell phones, smart phones, media phones, storage devices, portable media players, navigation systems, monitors, power supplies, adapters, remote control devices, chargers, and other devices. Encoded signals can be transmitted using interface circuits and connector receptacles that can provide pathways for signals and power compliant with various standards such as one of the Universal Serial Bus (USB) standards including USB Type-C, High-Definition Multimedia Interface® (HDMI), Digital Visual Interface (DVI), Ethernet, DisplayPort, Thunderbolt™ Lightning, Joint Test Action Group (JTAG), test-access-port (TAP), Directed Automated Random Testing (DART), universal asynchronous receiver/transmitters (UARTs), clock signals, power signals, and other types of standard, non-standard, and proprietary interfaces and combinations thereof that have been developed, are being developed, or will be developed in the future.

Various embodiments of the present invention can incorporate one or more of these and the other features described herein. A better understanding of the nature and advantages of the present invention can be gained by reference to the following detailed description and the accompanying drawings.

**BRIEF DESCRIPTION OF THE DRAWINGS**

**DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS**

**110** can be in communication with a second device **120** over a cable **130**. Specifically, connector insert **132** on cable **130** can be inserted into connector receptacle **112** on first device **110**, while a second connector insert (not shown) can be inserted into a second connector receptacle (not shown) on second device **120**. First device **110** and second device **120** can communicate by sending data to each other over cable **130**. One of the two devices can send power to the other over cable **130** as well.

This electronic device can include device circuitry **210**. Device circuitry **210** can include one or more integrated circuits, modules, or other circuits or compliments. Device circuitry **210** can implement some or all of the functions of the electronic device. The electronic device can further include transceiver circuitry. This transceiver circuitry can include error correction **220**. Error correction **220** can receive data from device circuitry **210** and can implement error correction algorithms and modify data to be transmitted accordingly. This circuitry can also, or instead, include other functions such as interleaving, run length limiting, or other functions. Encoder **230** can receive groups of bits and encode them for transmission by transmitter **240**. This encoder **230** can provide an efficient way of encoding data for transmission that reduces power, saves space, reduces design cycle time, and provides other benefits. The encoded data transmitted by transmitter **240** can be a multilevel code, for example, it can be PAM3 data. Transmitter **240** can transmit this data over data channel **250**. Data channel **250** can include connectors and cable **130** as shown in

Symbol data can be received by receiver **260** via data channel **250**. Receiver **260** can provide data to decoder **270**. This decoder **270** can provide an efficient way of decoding data for transmission that reduces power, saves space, reduces design cycle time, and provides other benefits. Decoder **270** can decode the symbols received from receiver **260** and provide groups of bits to error correction **280**. Error correction **280** can implement error correction algorithms and modify the received data accordingly. The circuitry can also or instead include other functions such as de-interleaving, length limiting, or other functions. Error correction **280** can provide data to device circuitry **210**.

In this specific embodiment, eleven bits can be encoded into a series of 7 three-level or ternary symbols. This coding can be selected for its efficiency since eleven bits defines 2048 possibilities, while 7 three-level symbols provides 2187 combinations into which the 2048 possibilities can be mapped. For example, binary bits can be encoded into symbols that can be pulse-amplitude modulated symbols or PAM3 symbols. In **310**. In act **320**, the bits can be arranged in branch bits, where a first branch BR**1** is equal to the first two bits, b**0** and b**1**, and where each subsequent branch is equal to the following three bits for four groups of branch bits. If the first branch bits have one of three possible values in act **330**, for example 00, 01, or 10, those two bits can be encoded as a first symbol in act **340**. The three subsequent groups of three bits each can then be encoded into two symbols in act **350**. The resulting 7 symbols can then be transmitted.

In these and other embodiments of the present invention, groups of three bits can be encoded into two three-level symbols. Three bits can have one of 8 values (8 different combinations), while two three-level symbols can have 9 possible values. This leaves a ninth state for the two symbols that is unused in the encoding of the three bits. This unused ninth state can then be used as an address value, where the presence and position of the address value can be used in decoding a series of symbols. In the translation table of **2**, b**3**, and b**4** (for example) can be encoded into PAM3 or other three-level symbols U**1** and U**2**, where U**1** and U**2** can have a low value, a zero value, or a high value. **1** and U**2** back into bits b**2**, b**3**, and b**4**. This table can be used for encoding various groups of three bits into two symbols, and back again, in the various examples shown below and by other embodiments of the present invention. In these and other embodiments of the present invention, other tables can be used. For example, a table having bits arranged in a gray-code order can be used.

The unused ninth state, in this example a value of HH for U**1** and U**2**, can be used as the address value in the various examples shown below and by other embodiments of the present invention, though in other embodiments of the present invention, other codes, such as LL or 00, can be used as the address value. This address value can be used to efficiently convey values of branch bits thereby simplifying encoding. These address values can also be used in a corresponding decoder to efficiently decode various branch bits, as shown below.

If the first branch bits from **2**, b**3**, and b**4**, which are referred to as the second branch bits BR**2**, are read and the encoding procedure continues at block A in **2**, the second branch bits BR**2** are encoded as the first symbol U**7** in acts **516**, **526**, and **536**, and one of the three subsequent groups of two symbols can be assigned the address value (HH) in acts **514**, **524**, and **534**. The first branch bits do not need to be encoded since their value (11) can be determined by a decoder from the presence of an address value. The remaining bits can be encoded into the remaining symbols as shown in acts **518**, **528**, and **538**.

In this specific embodiment of the present invention, 7 three-level symbols can be decoded into eleven bits. A group of seven symbols U**1**-U**7** can be received in act **610**. In this example, U**7** can be referred to as a first symbol and U**1**U**2**, U**3**U**4**, and U**5**U**6** can be referred to as symbol pairs. The symbol pairs can be checked for the presence of an address value in acts **612**, **622**, and **632**. If an address value is found, a value for the two first branch bits BR**1** can be determined to have a value of 11 in acts **614**, **624**, and **634**, consistent with the encoding scheme above. The first symbol can be read and from that the second branch bits BR**2** can be determined in acts **616**, **626**, and **636**. The remaining bits can be determined given the position of the address bits in acts **618**, **628**, and **638**. More specifically, since the first branch bits and the second branch bits are known, 6 other bits in the last two groups of three bits, remain to be determined. The address value can be located in a pair of the remaining 6 symbols, leaving two groups of two symbols to be decoded. Each group of two symbols can then be decoded into three bits, thereby completing the decoded set of eleven bits.

If an address value is not present in the 7 three-level symbols, the decoding can proceed to block A in **712**. The remaining three groups of two bits each can be decoded into three groups of three bits in act **714**, thereby completing the eleven decoded bits.

In a specific embodiment, 19 bits can be encoded into a series of 12 three-level symbols. This coding can be selected for its efficiency since 19 bits defines 524,288 possibilities, while 7 three-level symbols provides 531,441 combinations into which the 524,288 possibilities can be mapped. As before, these binary bits can be encoded into symbols that can be pulse-amplitude modulated symbols or PAM3 symbols. In **810**. The first of these bits, b**0**, can be used as the first branch bit BR**1** and subsequent groups of three bits can be referred to as subsequent branch bits in act **820**. If the first branch bit BR**1** has a specific value, for example 0 in act **830**, then the remaining 18 bits can be arranged in 6 groups of three bits each and encoded into a corresponding 6 pairs of two symbols for a total of 12 symbols in act **840**. This encoding can be the same or similar to the encoding in

If the first branch bit b**0** does not have the specific value, for example it is a 1, then a subsequent set of three bits b**1**, b**2**, and b**3**, can be used as second branch bits and the encoding can continue in **2** have a value in a second set of values in acts **912**, **922**, **932**, **1012**, **1022**, and **1032**, then one pair of symbols can be assigned the address value in acts **914**, **924**, **934**, **1014**, **1024**, and **1034**, where the location of the address value is dependent on the value of the second branch bits. The remaining 5 groups of three bits can be encoded into the remaining 5 pairs of symbols in acts **918**, **928**, **938**, **1018**, **1028**, and **1038**. There is no need to encode the first branch bit or the second branch bits since their value can be determined by a decoder from the presence of an address value.

If the second branch bits do not have a value in the second set of values, then a subsequent set of three bits can be used as third branch bits BR**3** and the encoding can continue in **1112**, and the third branch bits have a value in a third set of values in acts **1114**, **1124**, **1124**, **1214**, **1224**, and **1234**, then two pairs of symbols can be assigned the address value in acts **1116**, **1126**, **1136**, **1216**, **1226**, and **1236**, where the location of the address values are dependent on the value of the second branch bits and the third branch bits. The remaining 4 groups of three bits can be encoded into the remaining 4 pairs of symbols in acts **1118**, **1128**, **1138**, **1148**, **1218**, **1228**, **1238**, and **1248**. Again, there is no need to encode the first branch bit since its value can be determined by a decoder from the presence of an address value. There is also no need to encode the second or third branch bits since their value can be determined by a decoder from the presence of two address values.

If the second branch has a does not have the first specific value in act **1112**, the encoding can continue in **1314**, **1324**, **1324**, **1344**, **1414**, **1424**, and **1434**, **1444**, then two pairs of symbols can be assigned the address value in acts **1316**, **1326**, **1336**, **1346**, **1416**, **1426**, **1436**, and **1448**, where the location of the address values are dependent on the value of the second branch bits and the third branch bits. The remaining 4 groups of three bits can be encoded into the remaining 4 pairs of symbols in acts **1318**, **1328**, **1338**, **1348**, **1418**, **1428**, **1438**, and **1448**. Again, there is no need to encode the first branch bit since its value can be determined by a decoder from the presence of an address value. There is also no need to encode the second or third branch bits since their value can be determined by a decoder from the presence of two address values.

If the third branch bits do not have a value in the third set of values, then a subsequent set of three bits can be used as fourth branch bits. Three pairs of symbols can be assigned the address value in acts **1516**, **1526**, **1536**, **1546**, **1616**, **1626**, **1636**, and **1646**, where the location of the address values are dependent on the value of the second branch bits, the third branch bits, and the fourth branch bits. The remaining 3 groups of three bits can be encoded into the remaining 3 pairs of symbols in acts **1518**, **1528**, **1538**, **1548**, **1618**, **1628**, **1638**, and **1648**. Again, there is no need to encode the first branch bit since its value can be determined by the presence of an address value. There is also no need to encode the second, third, or fourth branch bits since their value can be determined by a decoder from the presence of three address values.

In a specific embodiment, a series of 12 three-level symbols can be decoded into 19 bits. In this example, the 12 symbols can be arranged in 6 pairs in act **1710**. Three of those 6 pairs can have the address value in one of 8 combinations. The 6 symbols pairs are checked to see if they have 3 pairs with address values in one of those 8 combinations in act **1712** in **1716**, **1726**, **1736**, **1746**, **1816**, **1826**, **1836**, and **1846**. The remaining bits can be decoded from the remaining symbols in acts **1718**, **1728**, **1738**, **1748**, **1818**, **1828**, **1838**, and **1848**.

If three symbol pair do not have the address value, a check of whether two symbol pairs have the address value is done in act **1912** on **1916**, **1926**, **1936**, **1946**, **2016**, **2026**, **2036**, **2046**, **2116**, **2126**, **2136**, **2146**, **2216**, **2226**, and **2236**. The remaining bits can be decoded from the remaining symbols in subsequent acts as shown.

If two symbol pairs do not have the address value, a check of whether one symbol pair has the address value is done in act **2312** and on **2510**. The remaining bits can be decoded from the remaining symbols. If no symbol pair have the address value, the first branch bit can be set in act **2510** and the symbol pairs can be decoded directly in act **2520** in

In these and other embodiments of the present invention, eleven bits can be encoded into a series of 7 three-level symbols. This coding can be selected for its efficiency since eleven bits defines 2048 possibilities, while 7 three-level symbols provides 2187 combinations into which the 2048 possibilities can be mapped. But this means that not all of the three-level states are used. This can lead to random data not having an average value of the middle state. The resulting baseline wander can reduce a usable signal amplitude and lead to errors. Accordingly, embodiments of the present invention can modify the above encryption schemes to reduce the resulting baseline wander. For example, in **2** can be encoded as one of three positions for the address value and the first symbol U**7**. U**7** can be L for three states, 0 for three states, and H for two states. As a result, the average value for U**7** might not be 0. Accordingly, when BR**2** has a value of 100, U**7** can be alternately set to H. The result can then be that U**7** can be L for three states, 0 for two states, and H for three states. As a result, the average value for U**7** can be 0. Other methods of reducing baseline wander are shown in the following figures.

In **2610**. In act **2620**, the bits can be arranged in branch bits, where a first branch BR**1** is equal to the first two bits, b**0** and b**1**, and where each subsequent branch is equal to the following three bits for four groups of branch bits. If the first branch bits have one of three possible values in act **2630**, for example 00, 01, or 10, those two bits can be encoded as a first symbol in act **2640**. The three subsequent groups of three bits each can then be encoded into two symbols in act **2650**. The resulting 7 symbols can then be transmitted.

If the first branch bits from **2**, b**3**, and b**4**, which are referred to as the second branch bits BR**2**, are read and the encoding procedure continues at block A in **2**, the second branch bits BR**2** are encoded as the first symbol U**7** in acts **2716**, **2726**, and **2736**, and one of the three subsequent groups of two symbols can be assigned the address value (HH) in acts **2714**, **2724**, and **2734**. The first branch bits do not need to be encoded since their value (11) can be determined by a decoder from the presence of an address value. The remaining bits can be encoded into the remaining symbols as shown in acts **2718**, **2728**, and **2738**.

In this specific embodiment of the present invention, 7 three-level symbols can be decoded into eleven bits. A group of seven symbols U**1**-U**7** can be received in act **2810**. In this example, U**7** can be referred to as a first symbol and U**1**U**2**, U**3**U**4**, and U**5**U**6** can be referred to as symbol pairs. The symbol pairs can be checked for the presence of an address value in acts **2812**, **2822**, and **2832**. If an address value is found, a value for the two first branch bits BR**1** can be determined to have a value of 11 in acts **2814**, **2824**, and **2834**, consistent with the encoding scheme above. The first symbol can be read and from that the second branch bits BR**2** can be determined in acts **2816**, **2826**, and **2836**. The remaining bits can be determined given the position of the address bits in acts **2818**, **2828**, and **2838**. More specifically, since the first branch bits and the second branch bits are known, 6 other bits in the last two groups of three bits, remain to be determined. The address value can be located in a pair of the remaining 6 symbols, leaving two groups of two symbols to be decoded. Each group of two symbols can then be decoded into three bits, thereby completing the decoded set of eleven bits.

If an address value is not present in the 7 three-level symbols, the decoding can proceed to block A in **2912**. The remaining three groups of two bits each can be decoded into three groups of three bits in act **2914**, thereby completing the eleven decoded bits.

These and other embodiments of the present invention can provide further techniques for reducing baseline wander. For example, the translation table of

**3000** can be translated to an offset value **3010**. In this example, eight different offset values are shown. In **3140**.

In these and other embodiments of the present invention can vary the above examples to further reduce baseline wander. An example is shown in the following figures.

In **3210**. In act **3220**, the bits can be arranged in branch bits, where a first branch BR**1** is equal to the first two bits, b**0** and b**1**, and where each subsequent branch is equal to the following three bits for four groups of branch bits. If the first branch bits have one of three possible values in act **3230**, for example 00, 01, or 10, those two bits can be encoded as a first symbol in act **3240**. The three subsequent groups of three bits each can then be encoded into two symbols in act **3250**. The resulting 7 symbols can then be transmitted.

As before, groups of three bits can be encoded into two three-level symbols. Three bits can have one of 8 values (8 different combinations), while two three-level symbols can have 9 possible values. This leaves a ninth state for the two symbols that is unused in the encoding of the three bits. This unused ninth state can then be used as an address value, where the presence and position of the address value can be used in decoding a series of symbols. In the translation table of **2**, b**3**, and b**4** (for example) can be encoded into PAM3 or other three-level symbols U**1** and U**2**, where U**1** and U**2** can have a low value, a zero value, or a high value. The translation table of **1** and U**2** back into bits b**2**, b**3**, and b**4**. This table can be used for encoding various groups of three bits into two symbols, and back again, in the various examples shown below and by other embodiments of the present invention. In these and other embodiments of the present invention, other tables can be used. For example, a table having bits arranged in a gray-code order can be used.

The unused ninth state, in this example a value of 00 for U**1** and U**2**, can be used as the address value in the various examples shown below and by other embodiments of the present invention, though in other embodiments of the present invention, other codes, such as LL or HH, can be used as the address value. This address value can be used to efficiently convey values of branch bits thereby simplifying encoding. These address values can also be used in a corresponding decoder to efficiently decode various branch bits, as shown below.

If the first branch bits from **2**, b**3**, and b**4**, which are referred to as the second branch bits BR**2**, are read and the encoding procedure continues at block A in **2**, the second branch bits BR**2** are encoded as the first symbol U**7** in acts **3416**, **3426**, and **3436**, and one of the three subsequent groups of two symbols can be assigned the address value (HH) in acts **3414**, **3424**, and **3434**. The first branch bits do not need to be encoded since their value (11) can be determined by a decoder from the presence of an address value. The remaining bits can be encoded into the remaining symbols as shown in acts **3418**, **3428**, and **3438**.

In this specific embodiment of the present invention, 7 three-level symbols can be decoded into eleven bits. A group of seven symbols U**1**-U**7** can be received in act **3510**. In this example, U**7** can be referred to as a first symbol and U**1**U**2**, U**3**U**4**, and U**5**U**6** can be referred to as symbol pairs. The symbol pairs can be checked for the presence of an address value in acts **3512**, **3522**, and **3532**. If an address value is found, a value for the two first branch bits BR**1** can be determined to have a value of 11 in acts **3514**, **3524**, and **3534**, consistent with the encoding scheme above. The first symbol can be read and from that the second branch bits BR**2** can be determined in acts **3516**, **3526**, and **3536**. The remaining bits can be determined given the position of the address bits in acts **3518**, **3528**, and **3538**. More specifically, since the first branch bits and the second branch bits are known, 6 other bits in the last two groups of three bits, remain to be determined. The address value can be located in a pair of the remaining 6 symbols, leaving two groups of two symbols to be decoded. Each group of two symbols can then be decoded into three bits, thereby completing the decoded set of eleven bits.

If an address value is not present in the 7 three-level symbols, the decoding can proceed to block A in **3612**. The remaining three groups of two bits each can be decoded into three groups of three bits in act **3614**, thereby completing the eleven decoded bits.

These and other embodiments of the present invention can utilize a translation table or similar technique, such as the translation table as shown in

These and other embodiments of the present invention can provide encoders and decoders that can be readily implemented using a minimal amount of logic gates. This can reduce component size, save power, speed design, and improve yields. While examples are shown utilizing specific numbers of bits, symbols, and type of symbols, these and other embodiments of the present invention can utilize different numbers of bits, symbols, and different types of symbol, such as four or five level symbols. Other types of encoding (and decoding), such as phase or frequency encoding, can also be used.

Embodiments of the present invention can provide data encoders and decoders that can be used in various types of devices, such as lighting, portable computing devices, tablet computers, desktop computers, laptops, all-in-one computers, wearable computing devices, cell phones, smart phones, media phones, storage devices, portable media players, navigation systems, monitors, power supplies, adapters, remote control devices, chargers, and other devices. Encoded signals can be transmitted using interface circuits and connector receptacles that can provide pathways for signals and power compliant with various standards such as one of the Universal Serial Bus standards including USB Type-C, High-Definition Multimedia Interface, Digital Visual Interface, Ethernet, DisplayPort, Thunderbolt, Lightning, Joint Test Action Group test-access-port, Directed Automated Random Testing, universal asynchronous receiver/transmitters, clock signals, power signals, and other types of standard, non-standard, and proprietary interfaces and combinations thereof that have been developed, are being developed, or will be developed in the future.

The above description of embodiments of the invention has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form described, and many modifications and variations are possible in light of the teaching above. The embodiments were chosen and described in order to best explain the principles of the invention and its practical applications to thereby enable others skilled in the art to best utilize the invention in various embodiments and with various modifications as are suited to the particular use contemplated. Thus, it will be appreciated that the invention is intended to cover all modifications and equivalents within the scope of the following claims.

## Claims

1. A method of encoding binary data, the method comprising:

- receiving a first number of bits including first branch bits and second branch bits;

- determining if the first branch bits have a value in a first set of values, and if they do, then

- encoding the first branch bits into a first symbol in a second number of symbols;

- encoding the remaining bits in the first number bits into the remaining symbols in the second number of symbols; and if the first branch bits do not have a value in the first set of values, then

- assigning an address value to two or more symbols in the second number of symbols; and

- not encoding the first branch bits; and

- encoding the remaining bits into the remaining symbols in the second number of symbols.

2. The method of claim 1 wherein the symbols are three-level symbols.

3. The method of claim 2 wherein the first number of bits comprises 11 bits, the first branch bits comprises 2 bits, and the second branch bits comprises three bits.

4. The method of claim 3 wherein the second number of symbols is 7 symbols.

5. The method of claim 4 wherein the address value is assigned to symbols based on the value of the second branch bits.

6. The method of claim 5 wherein the second branch bits are encoded as a first symbol and the address value is assigned to one of the three pairs of subsequent symbols.

7. The method of claim 6 wherein groups of three bits are encoded into pairs of symbols, such that 8 of 9 possible combinations for the two symbols are used, and the unused combination is the address value.

8. A method of decoding symbol data, the method comprising:

- receiving a second number of symbols, the symbols comprising a first symbol and a plurality of symbol pairs;

- determining if any of the plurality of symbol pairs has an address value, and if it does then

- setting first branch bits to a specific value;

- decoding the first symbol to second branch bits;

- decoding the remaining symbol pairs to a plurality of bits; and if none of the symbol pairs has the address value, then

- decoding the first symbol to the first branch bits; and

- decoding the remaining symbol pairs to a plurality of bits.

9. The method of claim 8 wherein the symbols are three-level symbols.

10. The method of claim 9 wherein the first number of bits comprises 11 bits, the first branch bits comprises 2 bits, and the second branch bits comprises three bits.

11. The method of claim 10 wherein the second number of symbols is 7 symbols.

12. The method of claim 11 wherein the symbols are decoded based on the position of the address value.

13. The method of claim 12 wherein before decoding, groups of three bits are encoded into pairs of symbols, such that 8 of 9 possible combinations for the two symbols are used, and the unused combination is the address value.

**Patent History**

**Publication number**: 20220376704

**Type:**Application

**Filed**: Aug 5, 2022

**Publication Date**: Nov 24, 2022

**Applicant**: Apple Inc. (Cupertino, CA)

**Inventors**: William P. Cornelius (Saratoga, CA), Seungyong Baek (San Jose, CA)

**Application Number**: 17/882,464

**Classifications**

**International Classification**: H03M 7/30 (20060101);