Security Issues - Peer-to-Peer Computing: Applications, Architecture, Protocols, and Challenges - page 167

Θ(N ) and the time taken to disinfect the system is Θ( N−2P N

P N ). For example,
for Code-Red worm, roughly 200,000 hosts will be infected in 7 hours and it
takes about 25 hours to clean up the system.
On the other hand, with the help of a P2P system, the maximum number
of infected hosts is Θ(I N (

P N ) γ ) and the time taken to disinfect the system is

N

+ γ

N
P N

Θ(ln

1+γ ln I N ). For example, for Code-Red worm, even with γ = 2, the
maximum number of infected hosts is on the order of 1000 only and it takes
only 5 hours to clean up the system.

7.7 P2P SIP
A key component in VoIP systems, the SIP (Session Initiation Protocol)
is also considered as an ideal candidate to be carried by a P2P network, from
the robustness perspective [Chopra et al., 2009, Seedorf, 2006]. Specifically,
instead of relying on centralized servers, a DHT is employed for registering
and locating a user ID (i.e., the SIP-URI). However, while the robustness
advantage of this approach is attractive, that is also associated with a whole
lot of security concerns. Indeed, most of the security problems that we discuss
above could render the P2P SIP protocol crippled. For example, the routing
table poisoning attack could make a peer unreachable or overwhelmed with
unnecessary tra c.

7.8 Collusive Piracy
Content piracy has always been a serious problem, even before the advent
of the Internet. The proliferated use of P2P computing clients only makes
this problem worse—many P2P users simply ignore copyright issues to share
commercial contents, reducing the potential profits of online content delivery.
Lou and Hwang [Lou and Hwang, 2009] proposed a very interesting ap-
proach to combat collusive piracy in a P2P network. Specifically, while content
poisoning, as described above, is considered as a vice rather than a virtue, Lou
and Hwang suggested a scheme to deliberately poison the paid contents when
the system detects that some pirates are downloading them.
In their proposed scheme, each peer is identified with its endpoint address,
consisting of an IP address and a listening port number. Notice that for peers
behind an NAT device, the public address representing the NAT box is used as
the IP address. Files to be shared are then incorporated with digital signatures
derived from the endpoint addresses. Legitimate clients can verify the digital