Check out the new USENIX Web site.
: Implications on Unwanted Traffic : Push vs. Pull: Implications : The Sender-Push Model

The Receiver-Pull Model

In the receiver-pull model, it is the receiver who initiates the message transfer by explicitly contacting the sender. The sender passively waits for the receiver and delivers the entire content upon receiving a request. Since it is the receiver who initiates the message transfer, the receiver would have explicit greater control over the message transfer and implicit greater trust in the received content, than in the sender-push model.

A number of successful communication services rely on the receiver-pull model. The most important examples using the receiver-pull model are the FTP and HTTP protocols. In both cases, the receiver initiates the data transfer by opening an FTP connection or by typing/clicking on a URL, respectively. (Interestingly, HTTP supports both receiver-pull and as well as RISP variant of sender-push, though the former is more commonly used. Examples of RISP model techniques in HTTP include automatic page refreshes and the hugely unpopular popup windows).

An interesting and useful variation of receiver-pull model, which is of special interest to us, is the sender-intent-based-receiver-pull (SIRP). In this model, the sender first expresses an intent to send content to the receiver via a small intention message. If the receiver happens to be interested, it contacts the sender and retrieves the content. A common example of the SIRP model is the pager service. Here the caller expresses an intent to talk to a callee by paging the latter and leaving a callback number. If the callee is interested, he/she contacts the caller back on the callback number. The main feature of the SIRP model is that the content itself is pulled by the receiver whereas only a short intent is pushed by the sender.

The primary advantage of the receiver-pull model is that a receiver exercises control over when and what it receives. The receiver has the freedom to first determine its own level of interest in the content (as well as the reputation of the sender) before it actually requests the content. Secondly, it becomes the responsibility of the sender to store and manage the content till the receiver is ready to retrieve it. For instance, an FTP or web server needs to store and manage its own files whereas receivers access it only when they are interested. Thirdly, there is a large window of time over which a malicious sender is forced to stay online and reveal its identity. For the pure receiver-pull model, this window is from the moment content is generated and named till the content is retrieved by the receiver. For the SIRP model, this window is from the moment sender expresses its intent to send till the time receiver retrieves the content. Thus, unlike the sender push model, there is a large window of time in which the receiver is free to verify a sender's identity.

One obvious disadvantage of receiver-pull model is that the sender is burdened with greater content management complexity. The sender needs to store outgoing messages and keep them available at least till the intended receivers are willing to retrieve them, and needs to have a deletion policy if a message is never retrieved by the receiver. Another issue that the sender needs to grapple with is to ensure that the party retrieving a message is indeed the originally intended receiver. However, another angle to look at these disadvantages is that, in the sender-push model, it is the receiver who needs to deal with the very same issues.

: Implications on Unwanted Traffic : Push vs. Pull: Implications : The Sender-Push Model