Make the most of the unlicensed ISM band

The ISM band is an unlicensed band intended for low-data rate communications and extends from 902 to 928 MHz. By definition, any transceiver operating in an unlicensed band doesn''t require that the end user obtain a government permit to use the device. However, the device itself must be certified by the governing authority in the operating country. In the U.S., low-data rate transceivers operating in the 915-MHz band are governed by the FCC, whose regulations focus exclusively on these devices'' emissions.

Spread-spectrum systems are known for their interference rejection and anti-jamming capability in multiple access systems. Spread-spectrum signals used to transmit digital information are distinguished by the characteristic that their bandwidth, W, is much greater than the information rate, R, in bits/s. The bandwidth expansion factor, Be = W/R for a spread spectrum signal, is much greater than unity. The large redundancy inherent in spread-spectrum signals is needed to overcome the potentially severe interference levels that are encountered when transmitting digital information over radio channels. Radios operating in the 915-MHz ISM band are subjected to interference from other unlicensed devices in the same band which could significantly degrade the operating radio''s sensitivity.

Another important element employed in the design of spread-spectrum systems is pseudo-randomness, which makes the transmitted signals appear similar to random noise and difficult to demodulate by the receivers, other than the ones intended to receive the information. To summarize, the benefits of spread spectrum systems are:

• combating or suppressing the interference''s detrimental effects due to jamming, interference arising from other users of the same channel and self-interference due to multipath propagation;
• achieving message privacy in the presence of other listeners;
• hiding a signal by transmitting it at low power and thus making it difficult for an unintended listener to detect in the presence of background noise. Thus, these signals have a low probability of intercept (LPI).

Two types of spread-spectrum systems exist: direct-sequence (DSSS) and frequency-hopping (FHSS). In DSSS, a pseudo-noise (PN) generator generates PN codes at the chip rate, which is much faster than the data rate. The data output at the data rate and the PN generator output at the chip rate are Modulo-2 added and fed to the phase-shift-keying (PSK) modulator. At the receiver, the PN codes'' complex correlation properties decode the message sequence. DSSS is an expensive solution and more complex to implement due to the stringent synchronization requirements. DSSS also needs a coherent modulation technique like binary PSK. These factors exclude DSSS as a suitable choice for simple, low-cost, low-data rate ISM band transceivers. The alternative, FHSS, is better suited for such applications.

FHSS systems
In an FHSS system, the available channel bandwidth is subdivided into a large number of contiguous frequency slots. The US ISM band lends itself to low data rate FHSS systems. In any signaling interval, the transmitted signal occupies one of the available frequency slots. Selecting the frequency slots during each signaling interval is made pseudo-randomly. User-defined protocols can determine the hopping sequence (Fig. 1).

The time the transceiver is transmitting or receiving is called the dwell time. The time when the transceiver is configuring its registers to transmit or receive at another frequency is called the blank time. What occurs during dwell time is the transmit (or receive) preamble, start bit, data sequence, post-amble at a particular frequency in the hopping sequence. And the actions during blank time are pseudo-random frequency generation, configuring the transceiver registers to operate at the randomly generated frequency, and waiting for the PLL to lock. To conserve battery power, the blank time should be as short as possible.

There''s a good reason to use frequency hopping for transceivers operating in the ISM band. FCC regulates the operation of unlicensed devices in this band. Under part 15.247 of the FCC regulations, frequency-hopping systems are permitted to transmit at powers of up to +30 dBm EIRP. This higher power operation (and hence higher link budget or, ultimately, range) coupled with the benefits of spread spectrum systems makes frequency hopping an attractive option for the unlicensed radio devices in the ISM band (see the table).

Single-channel general-purpose data devices share the same 902- to 928-MHz frequency band as hoppers, but operate at reduced power levels. There''s no hopping requirement for the single-channel devices. The maximum output power is about "1.25 dBm, and the maximum harmonic power is about "41.25 dBm.

Frequency-hopping UART chip set
One frequency hopping wireless UART chip set takes advantage of interference-rejection features and the higher transmit power option offered by employing frequency hopping systems. The chip set, operates in the 902- to 928-MHz ISM band, is a true wireless UART solution (data-in RF-out, RF-in data-out) where all the hardware and software aspects of a wireless link have already been implemented to yield a ready-made solution. The chip set can implement a wireless link that end applications can interface to as a peripheral, shielding the application from the frequency hopping system''s complex implementation details.

The chip set''s transceiver is identified by a 16-bit unique transceiver ID, a 16-bit network ID, and a 16-bit system ID, forming a 48-bit address. Each transceiver is also configured to use one of 16 distinct frequency hop-sets, which consists of 50 frequency channels pseudo-randomly arranged. Two transceivers with different hop-set configurations will not communicate.

The data packet structure consists of a header, the data payload, and the trailer with the checksum. The data communication protocol supports acknowledgments and retries (up to 20) for ultra-reliable data transfer. The chip set is configured to support point-to-point and point-to-multipoint (broadcast) network topologies.

Frequency-hopping protocol
Each transceiver is designed to hop pseudo-randomly across 50 channels in the 902- to 928-MHz band as configured by the hop-set ID. Because the transmit carrier frequency hops pseudo-randomly, the receiver needs to generate frequencies in the same pseudo-random order to ensure proper frequency lock, demodulation, and detection of the signal. Thus, there has to be time synchronization between the transmitter and receiver. This synchronization happens in two phases, acquisition and tracking. Acquisition is the initial phase where the receiver recognizes the transmitter. The tracking phase happens upon successful acquisition. While tracking, the transmitter and the receiver need to be in continuous synchronization until data transmission and/or reception is complete. Both phases can be implemented in firmware.

The transmit and receive devices must be set to identical hop tables. The originating device, once activated to transmit, transmits data on a random channel determined by the configured hop set. The receive device scans each channel looking for the TX preamble consisting of a 0101 sequence. Once the receive device determines a valid preamble, it remains on the valid channel. When the originating device transmits the 70-ms preamble it sends the sync pattern (with the sequence of 00110011), the receive device syncs up with the originating device and prepares to receive valid data. Upon receiving valid data, the receiver hops to the next channel predetermined by firmware to transmit an acknowledgement to the originating device. The device then goes into receive mode after transmission and listens for the acknowledgement on the next channel determined by firmware. Upon successful communication, the originating device passes a successful transmission acknowledgement from the intended receiver to the host (Fig. 2).

Because the chip set can implement a pseudo-random frequency hopping protocol, there''s considerable interference rejection offered when compared to transmitting data on a single-channel. Using a frequency-hopping system for the 915-MHz ISM band has several advantages:

• Good rejection against narrow-band and broadband interference.
• Less expensive and simpler implementation when compared to a DSSS modulation.
• Excellent performance in indoor multi-path environment.
• Well suited for low-data rate systems because higher packet overhead isn''t a major concern.
• Low-probability of intercept because frequency-hop signals have a low average power density which can make these signals difficult to intercept. While the frequency''s instantaneous power level that''s transmitted is high, the average power of that frequency is equal to the instantaneous power divided by the number of frequency slots.
• Short acquisition/synchronization time when compared to DSSS.
• Frequency hop systems provide a great degree of message privacy. The low intercept
• probability combined with pseudorandom frequency hopping make these signals difficult to demodulate for unintended receivers.
• Simple wireless UART with all aspects of the frequency hopping and the associated communication protocols implemented in ROM.

About the author
ShreHarsha Rao is an RF systems and applications engineer at Texas Instruments in the high-speed and RF division. He received his BSEE from Bangalore University and MSEE from the University of Texas at Arlington. He can be reached at harsha@ti.com.