COMPUTATIONAL TASKS IN MEDICAL NANOROBOTICS - Bio-Inspired and Nanoscale Integrated Computing - page 421

the nanorobot to engage in actions that will safely remove it from the body since it
can no longer fulfill its mission. One simple failsafe mechanism widely employed in
high-reliability programmable devices is the watchdog timer—a counter that can
shut down the computer if it ever reaches zero, but which is continually reset by a
correctly operating program so that it never reaches zero as long as the program
continues functioning (analogous to the biological ''apoptotic switch'' described
earlier). In the case of medical nanorobots that require active supervision, the reset
command could be periodically rebroadcast to the nanodevices in vivo by the
attending physician via ultrasound messaging.
Safety protocols may range from relatively simple procedures, such as self-
diagnostic routines, to very complex procedures such as dental protocols which
instruct mobile nanorobots bled into the mouth to avoid the hard grinding
surfaces by retreating to lower positions on the teeth [1z], or flight protocols for
aerial nanorobots including self-enforcement of no-fly zones near nose and mouth
along with other active anti-inhalation, inhale-safe, and post-inhalation extraction
protocols [1bp]. Other safety protocols may incorporate a wide variety of user-set
locks and limits, command blocking, limits on access to functions by non-
physicians, and perhaps some equivalent to the humorously termed ''shame
blocker'' (a recent telephone gimmick in which the user dials 333 and a number,
and then his phone won't let him call that number, e.g., his ex-wife when he's
drunk).

15.4.5. Security Protocols
Even the best security cannot prevent all harm but can help avoid significant harm
while allowing the system to continue operating normally. While most features are
useful for what they do, security features within products are useful because of
what they don't allow to be done. Security engineering helps to ensure that the
nanodevice will not fail ''in the presence of an intelligent and malicious adversary
who forces faults at the worst time and in the worst way.'' Security protocols are
required to ensure that incoming commands originate from trusted and author-
ized sources, and may employ checksums, signed and certified programs, and
formally proofed systems without trapdoors. These protocols should normally
refuse to accept commands that could cause the device to exhibit behaviors that
would harm the patient. Communication protocols might include TCP/IP proto-
col stacks combined with typical security technologies such as firewalls, packet
filtering, intrusion detection, and secure procedures for flow control and authen-
tication (e.g., passwords, biometrics, and public-key cryptography) to prevent
''body-hacking.'' Depending on circumstances, nanorobots may need blockers for
viruses, worms, and spam, and protocols linked to sensors that monitor structural
integrity to ensure tamper resistance.
Other security protocols may be required to authorize reading and writing
data into personal medical record caches implanted in the patient's body—
multiply redundant caches 1mm 3 1000 TB of fast-access

mechanical memory, capacious enough to store a lifetime of detailed medical data,

in size could hold W

B