17 November 2014

XSL - the "Japanese Slot Machine"

Japanese Slot Machine (also named XSL or JSM) is a system thought to be from the Japanese Government or Self-Defense Force (Navy). Some have likened its weird sound to that of a Las Vegas slot machine, so the name Japanese Slot Machine.
I heard an XSL transmission on 16 November (at 2212z) on 8588.0 KHz/USB: see below a screenshot from my SDR Console during the reception.

Known Frequencies
4231.5 kHz     6417.0 kHz     8588.0 kHz        
4291.0 kHz     6445.1 kHz     8704.0 kHz

The signal transmits continuously on pairs of frequencies in the 4MHz, 6MHz and 8MHz bands, which places it firmly in the ITU bands allocated for Maritime use.
Reports show the signal to be stongest in the Far East, indicating an origin in that part of the World. Although the signals are weak in Europe, they can be monitored in the evenings on both the 6MHz and 8MHz frequncies. The poor reception makes analysis of the signal difficult.
An article in Monitoring Times, December 2002 was the first to identify these signals as Japanese Navy. Writing it the "Utility World" column, Hugh Stegman outlines his reasons for this claim. Firstly, direction-finding fixes indicated Japan at the source of the signals, although China and Russia were not ruled out.
Secondly, the frequencies correlate with those previously used by the Japanese Navy for eight-tone radio modems, some of which disappeared at the right time.
Finally, monitors travelling to Japan identified additional strong local frequencies, some of which were only operating on a part daily basis.

The mode being used is quadrature phase-shift keying (QPSK) , encrypted shore to ships traffic.  The mode is un-decodable: Sigmira doesn't really decode the information but only displays the "frames" and raw QPSK symbols.

It still remains a challenging and difficult signal to monitor at any reasonable strength in Europe.

The following is quoted from the Sigmira manual.
"With some investigation the symbol rate was determined to be 1600.00 baud. The regular ticking sound was found to be an exactly repeating sequence of symbols. Clearly that serves as a channel probe and frame synchronization pattern. 
One tick sound period is here defined as a "frame". Frames were found to consist of 140 QPSK symbols. So the frame rate is 11.42857 Hz. 
The probe/sync pattern is 28 symbols which is one fifth the number of symbols per frame. It is found that, during the repetitive melody idle time, the remaining symbols of a frame consist of four repetitions of another 28 symbol pattern. So a frame appears to consist of five "blocks" of 28 symbols. 

During idle time there is a finite set of symbol patterns that appear in the blocks. The patterns are designated here: ps, p0, p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, and p11. The "ps" pattern is the probe/sync pattern. The rest are numbered roughly in the order of their frequency of occurrence. The "p0" pattern is most common. The melodic idle time consists of an exactly repeating sequence of 64 frames. A 64 frame sequence is called a "super frame" here. 

The duration of a super frame is 5.60000 seconds. The frame / block pattern of the idle super frame is presented in the following table. The melody arises from the regular repeated simple patterns of symbols/phases. The p10 pattern is all one phase. So it produces a single tone. The p11 pattern is also all one phase but 180 degrees from p10. The other patterns are somewhat more complex and result in different and multiple apparent tones."
A short recording of my XSL reception may be heard here
decoder Sigmira may be downloaded from the sigmira official site 

No comments:

Post a Comment