As much as we wax lyrical on the Internet being without borders and transcending governments, it's sobering--and important--to see just how quickly and easily a government can enforce its will.
Any networked system has a large number of vulnerabilities:
1. Physical hardware. You combat this in three possible ways:
a) Make the hardware as cheap as possible;
b) Make the hardware dual-purpose. For example, behind the Iron Curtain, they would modify radios beyond what the state allowed them to listen to; and
c) Use hardware that the state can not do without. For example, use of the telephone network by spies in occupied Europe during WW2.
2. Wired networks are inflexible and easily interrupted. It wouldn't take too many acts of sabotage to almost completely disable the Internet in the US. Wireless networks are subject to state tracking. You combat this in a number of ways:
a) Use a medium the state can't do without (eg power lines, although in extreme circumstances the state can and will cut the power);
b) Hide your communication channel (eg steganography);
c) Use a wireless medium where the transmitters are small, cheap and portable.
Note: in this context, carrying messages using people constitutes a wireless packet-switched network with high bandwidth and high latency, which solves the problems of cost and partially solves the problem of tracing as the state can still use armed forces to limit movement.
It's worth discussing encryption in particular. It is not a sufficient protection but it can help. For one thing, an intercepted encrypted message will be taken as a sign of guilt in extreme circumstances regardless of its content.
Also, the distribution of encryption keys is an age old problem.
Lastly, encryption has a human element. You don't really know who is receiving your message and what they're doing with it. They could be a government agent, a turncoat or simply surrender their secrets under duress.
So I'm not sure this is a problem that can be solved but it can be easier.
Any networked system has a large number of vulnerabilities:
1. Physical hardware. You combat this in three possible ways:
a) Make the hardware as cheap as possible;
b) Make the hardware dual-purpose. For example, behind the Iron Curtain, they would modify radios beyond what the state allowed them to listen to; and
c) Use hardware that the state can not do without. For example, use of the telephone network by spies in occupied Europe during WW2.
2. Wired networks are inflexible and easily interrupted. It wouldn't take too many acts of sabotage to almost completely disable the Internet in the US. Wireless networks are subject to state tracking. You combat this in a number of ways:
a) Use a medium the state can't do without (eg power lines, although in extreme circumstances the state can and will cut the power);
b) Hide your communication channel (eg steganography);
c) Use a wireless medium where the transmitters are small, cheap and portable.
Note: in this context, carrying messages using people constitutes a wireless packet-switched network with high bandwidth and high latency, which solves the problems of cost and partially solves the problem of tracing as the state can still use armed forces to limit movement.
It's worth discussing encryption in particular. It is not a sufficient protection but it can help. For one thing, an intercepted encrypted message will be taken as a sign of guilt in extreme circumstances regardless of its content.
Also, the distribution of encryption keys is an age old problem.
Lastly, encryption has a human element. You don't really know who is receiving your message and what they're doing with it. They could be a government agent, a turncoat or simply surrender their secrets under duress.
So I'm not sure this is a problem that can be solved but it can be easier.