48
edits
Changes
Jump to navigation
Jump to search
Line 38:
Line 38: − + − The packet header is fixed and consists in a synchronization byte (0xA5), followed by a unused (possibly RFU) zero byte. After these two hardcoded bytes, there's a varint representing the payload size, which may use one byte or two, depending on the how big the payload is.+ − − For packets with less than 64 bytes:+ − + − + − + − ! Size+ + + + + + + + + + + + + + + + + + + − + − + − + + + + + + + + − For packets with up to 16383 bytes:+ − + − + − + − ! Size (1)+ − + + + + + + + + + + + + + + + + + + + + + + + + + − + − + − + − + + + + + + + Line 87:
Line 141: − + Line 118:
Line 172: − +
no edit summary
Packets are sent using IrDA-SIR (using [[ir:USER]]), with a 8N1 encoding (eight data bits, one stop bit, without parity). Each one is formed by a 2-byte header, a varint with the payload size, an obfuscated payload, and trailing error detection byte.
Packets are sent using IrDA-SIR (using [[ir:USER]]), with a 8N1 encoding (eight data bits, one stop bit, without parity). Each one is formed by a 2-byte header, a varint with the payload size, an obfuscated payload, and trailing error detection byte.
== Packet header ==
== Layer 1 - framing format ==
Frames are encoded using two different yet very simmilar formats, depending on how large the payload to be transmitted is:
* For payloads with less than 64 bytes, the third byte represents the payload size.
* For payloads with less than 64 bytes, the third byte represents the payload size.
* For packets with up to 16383 bytes, the size is split in two bytes, with the third byte being the upper 6 bits of the payload size, OR'd with 0x40, and the fourth being the lower eight bits of the payload size
* For packets with up to 16383 bytes, the size is split in two bytes, with the third byte being the upper 6 bits of the payload size, OR'd with 0x40, and the fourth being the lower eight bits of the payload size
{| class="wikitable" style="margin: 0 auto; text-align: center;"
{| class="wikitable" border="1"
|+IR framing format - short frame
! Sync
|-
! RFU
! style="border-top: none" | Byte
! style="width:10%;"| 7
! style="width:10%;"| 6
! style="width:10%;"| 5
! style="width:10%;"| 4
! style="width:10%;"| 3
! style="width:10%;"| 2
! style="width:10%;"| 1
! style="width:10%;"| 0
|-
! 0x00
| colspan="8"|Synchronization byte (<code>0xA5</code>)
|-
! 0x01
| colspan="8"|Reserved for future use (<code>0x00</code>)
|-
! 0x02
| RFU (<code>0</code>)
| Short frame (<code>0</code>)
| colspan="6"|Payload size
|-
|-
| 0xA5
! 0x03
| 0x00
| colspan="8"|Payload byte 0
| size
|-
| colspan="9"|...
|-
! 0x03+n-1
| colspan="8"|Payload byte n-1
|-
! 0x03+n
| colspan="8"|[[CRC-8-CCITT]] computer over whole packet
|}
|}
{| class="wikitable" border="1"
{| class="wikitable" style="margin: 0 auto; text-align: center;"
! Sync
|+IR framing format - long frame
! RFU
|-
! style="border-top: none" | Byte
! Size (2)
! style="width:10%;"| 7
! style="width:10%;"| 6
! style="width:10%;"| 5
! style="width:10%;"| 4
! style="width:10%;"| 3
! style="width:10%;"| 2
! style="width:10%;"| 1
! style="width:10%;"| 0
|-
! 0x00
| colspan="8"|Synchronization byte (<code>0xA5</code>)
|-
! 0x01
| colspan="8"|Reserved for future use (<code>0x00</code>)
|-
! 0x02
| RFU (<code>0</code>)
| Long frame (<code>1</code>)
| colspan="6"|Payload size (upper 6 bits)
|-
! 0x03
| colspan="8"|Payload size (lower 8 bits)
|-
! 0x04
| colspan="8"|Payload byte 0
|-
|-
| 0xA5
| colspan="9"|...
| 0x00
|-
| (size >> 8) <nowiki>|</nowiki> 0x40
! 0x04+n-1
| size & 0xFF
| colspan="8"|Payload byte n-1
|-
! 0x04+n
| colspan="8"|[[CRC-8-CCITT]] computer over whole packet
|}
|}
=== Header ===
The packet header is fixed and consists in a synchronization byte (0xA5), followed by a unused (possibly RFU) zero byte. After these two hardcoded bytes, there's a varint representing the payload size, which may use one byte or two, depending on the how big the payload is.
In C:
In C:
}</nowiki>
}</nowiki>
== Payload ==
=== Payload ===
The payload is obfuscated using a XOR-based encryption. In C:
The payload is obfuscated using a XOR-based encryption. In C:
<nowiki>void payloadObfuscate(const void * voidplain, void * voidcipher, size_t size) {
<nowiki>void payloadObfuscate(const void * voidplain, void * voidcipher, size_t size) {
}</nowiki>
}</nowiki>
== Error detection ==
=== Error detection ===
The trailing error detection byte is calculated using [[CRC-8-CCITT]] <b>over the whole packet</b> (both the header and the payload)
The trailing error detection byte is calculated using [[CRC-8-CCITT]] <b>over the whole packet</b> (both the header and the payload)