Beyond Shannon-Weaver communication

Figure 10.13: The classical Shannon-Weaver model of communication (from 1948). The sender provides a message to the encoder, which transmits the message through a channel corrupted by noise. At the other end, a decoder converts the message into a suitable format for the receiver. This model serves as the basis of communication theory in engineering.

An important factor is how many people will be interacting through the medium. Start with a pair of people. One of the most powerful mathematical models ever developed is the Shannon-Weaver model of communication, which for decades has been the basis of design for communication systems in engineering; see Figure 10.13. The model involves a sender and a recipient. The communication system encodes a message from the sender, which is then sent over a noisy channel. At the other end, the system decodes the message and it arrives to the recipient. The recipient could give feedback to indicate whether the message has been received intact. This communication model gave rise to the field of information theory, which enabled a well-defined notion of bandwidth for a communication channel and revealed the limits of data compression.

This model is powerful in that it mathematically quantifies human interaction, but it is also inadequate for covering the kinds of interactions that are possible in VR. By once again following the universal simulation principle, any kind of human interaction that exists in the real world could be brought into VR. The Shannon-Weaver model is inspired by interaction mechanisms such as the 19th century telegraph or 20th century handheld receiver (or walkie-talkie). In these cases, the humans are completely isolated from each other, and the technology provides a burst of information that is similar to writing a letter. We have gone from text to audio to video communication, and could extend even further by incorporating displays for other senses, such as touch and smell. There are also so many opportunities to use synthetic models, possibly in combination with actual captured information from cameras and microphones. Simple gestures and mannerisms can provide subtle but important components of interaction that are not captured by the classical communication model.

In spite of its shortcomings for VR, keep in mind that the Shannon-Weaver model provides powerful analysis of bandwidth and latency for computer networks and systems, which ultimately support any form of social interaction. Therefore, it has far reaching implications on what can or cannot be accomplished in a VR system. This occurs because all ``communication'' is converted into streams of bits that are sent through cables or network connections. One key problem is to ensure that the targeted social interaction VR experience is comfortable, convincing, and reliably supported over the computer network.