previous | contents | next

Chapter 25 ½ ALOHA Packet Broadcasting: A Retrospect 425

problem of received phase ambiguity.) The resultant configuration is shown in Fig. 5.

Radio Range

The maximum operating distance between any terminal of the ALOHANET and the MENEHUNE (or a repeater) is specified as the system's radio range. This distance is primarily a function of a transmitter's radiated power, the receiver's sensitivity, and the attenuation of radio signal power for the given distance. Local noise conditions at the receiver location can also affect this distance, but for system planning purposes, range is usually calculated on the basis of some given propagation model. For line-of-sight paths, which exist at VHF, UHF, and higher frequencies, two different models are used depending upon local topographical conditions. In an urban area these paths are partially obstructed and suffer from multipath effects. A power loss proportional to I/R4 is usually assumed for these conditions [Okumura et al., 1968]. Where paths are unobstructed and well clear of the local terrain, a spreading loss proportional to I/R2 can be assumed. Receiver threshold sensitivity in the ALOHANET is defined as that receiver input power level which causes an average bit error rate of 10-5. This bit error rate should provide a packet throughput reliability better than 99 percent for full-length ALOHA packets.

Assuming a transmitter equivalent radiated power of 10 watts, a simple whip antenna at a user terminal, an elevated antenna at the MENEHUNE or repeater and a 3 microvolt receiver sensitivity, the radio range works out to about 17 miles in the urban area for the ALOHANET frequencies. Between repeaters and the MENEHUNE terminal, which have well-elevated antennae and good path clearances, the assumed I/R2 model gives a maximum range of 290 miles. The use of high-gain omnidirectional antenna arrays at repeater sites extends these ranges. Tests conducted on a 100 mile path between two ALOHANET repeaters confirmed the I/R2

spreading-loss assumption and indicated a fade margin of 30 db existed (due to the 10 db gain antennae used for the test.)

Data Synchronization

Because of the burst nature of radio transmission of ALOHANET packets, special synchronization techniques must be employed in the modem and data terminal equipment. Since the phase-shift-keying used in the ALOHANET modem design is a bit-synchronous technique, bit synchronization must first be performed in the demodulator before packet synchronization can be attempted. Bit-sync is performed by a phase-locking circuit, and a lock-indication signal is passed to the data equipment when bit-sync has been attained. The bit-sync detection circuit is so designed to provide a very low false detection probability (less than 10-6) and a high probability of packet detection. The narrow bandwidth of the phase-lock circuit presently designed into the ALOHANET modem requires a bit-sync preamble of 90 bits to ensure reliable bit-sync. Studies have indicated that this preamble can be reduced to about 10 bits by use of a redesigned wide-band phase-lock circuit. In fact, we are presently contemplating doing away with the bit-sync preamble entirely, further reducing packet overhead. The unique characteristics of the ALOHA modem design male such an approach feasible.

Packet synchronization is accomplished in the ALOHANET data terminal buffer by means of the 16-bit parity word contained in the packet header. When the parity check routine accepts the header, the packet is assumed to be synchronized. Since the parity check routine is initiated by the first bit of the header, packets can be missed due to detection of an early error bit before the header. This miss probability is presently controlled by the modem at about 10-3 or less, providing a packet detection probability of 99.9 percent or better. The false detection probability of this circuit is ~1.5´ 10-5, which is independent of that of the modem. Thus, the overall probability of false detection is less than 1.5´ 10-11. Therefore, less than one out of a thousand packets will be lost due to packet sync errors and packet sync false alarms occur with extreme rarity.
 
 

User Interface Choices

The development of the ALOHANET user interface has been an evolutionary process, as is typical of most research developments. Since there were expected to be many user nodes (as compared to the single MENEHUNE node), the primary design goals were initially set as simplicity of design and low cost. This led to the design of a hardwired control unit with limited data storage capability coupled to a modem and radio transceiver. This initial design was termed a Terminal Control Unit (TCU). As experience
 
 

previous | contents | next