Rattle quoted WSJ: President Bush has promised a frugal budget proposal next month, but one big-ticket item is stirring controversy: an estimated $6 billion to build a secretive system protecting U.S. communication networks from attacks by terrorists, spies and hackers.
Then Decius asked: Could it be related to this?
And by that you mean The Spymaster, which I recommended earlier this month. The article recommended by Rattle is here in full text. Significantly, the figure cited above is only the starting point: The administration’s plan is to reduce points of access between the Internet and the government and to use sensors to detect intrusions displaying potentially nefarious patterns, said former top intelligence officials. The program would first be used on government networks and then adapted to private networks. Former officials said the final price tag is approaching an estimated $30 billion over seven years, including a 2009 infusion of around $6 billion, though those numbers could change significantly as the plan develops.
This Chertoff quote is either amusing or disturbing, depending on your perspective: "There is a lot of thought being given to: How do you organize this in a way that protects an incredibly valuable asset in the United States but does it in a way that doesn’t alarm reasonable people, and I underline reasonable people, in terms of civil liberties?"
Finally: The CIA and the Pentagon didn’t want other agencies mucking about ...
This tussle is referred to at the end of the Washington Post coverage just now recommended here. What's silly here is that no one is talking about ROC curves. How can you even propose to monitor the open Internet? The human resources involved would be outrageous, no? Not quite as bad as having human telephone switch operators, but as presented, this proposal simply doesn't scale, and as such is not credible. The stated intention to "protect US networks from hackers" is not credible, because the proposed task cannot be resourced. How much can they really accomplish, anyway? Consider the following: Insertion, Evasion, and Denial of Service: Eluding Network Intrusion Detection All currently available network intrusion detection (ID) systems rely upon a mechanism of data collection -- passive protocol analysis -- which is fundamentally flawed.
Maybe they intend to install normalizers at every access router in the US? Network Intrusion Detection: Evasion, Traffic Normalization, and End-to-End Protocol Semantics A fundamental problem for network intrusion detection systems is the ability of a skilled attacker to evade detection by exploiting ambiguities in the traffic stream as seen by the monitor. We discuss the viability of addressing this problem by introducing a new network forwarding element called a traffic normalizer. The normalizer sits directly in the path of traffic into a site and patches up the packet stream to eliminate potential ambiguities before the traffic is seen by the monitor, removing evasion opportunities.
Of course even then you face The Eavesdropper's Dilemma. RE: WSJ | Bush Looks to Beef Up Protection Against Cyberattacks | 
