(* Title: HOL/Auth/NSP_Bad
ID: $Id$
Author: Lawrence C Paulson, Cambridge University Computer Laboratory
Copyright 1996 University of Cambridge
Inductive relation "ns_public" for the Needham-Schroeder Public-Key protocol.
Flawed version, vulnerable to Lowe's attack.
From page 260 of
Burrows, Abadi and Needham. A Logic of Authentication.
Proc. Royal Soc. 426 (1989)
*)
AddEs spies_partsEs;
AddDs [impOfSubs analz_subset_parts];
AddDs [impOfSubs Fake_parts_insert];
AddIffs [Spy_in_bad];
(*For other theories, e.g. Mutex and Lift, using AddIffs slows proofs down.
Here, it facilitates re-use of the Auth proofs.*)
AddIffs (map simp_of_act [Fake_def, NS1_def, NS2_def, NS3_def]);
Addsimps [Nprg_def RS def_prg_simps];
(*A "possibility property": there are traces that reach the end*)
Goal "A ~= B ==> EX NB. EX s: reachable Nprg. \
\ Says A B (Crypt (pubK B) (Nonce NB)) : set s";
by (REPEAT (resolve_tac [exI,bexI] 1));
by (res_inst_tac [("act", "NS3")] reachable.Acts 2);
by (res_inst_tac [("act", "NS2")] reachable.Acts 3);
by (res_inst_tac [("act", "NS1")] reachable.Acts 4);
by (rtac reachable.Init 5);
by (ALLGOALS (asm_simp_tac (simpset() addsimps [Nprg_def])));
by (REPEAT_FIRST (rtac exI ));
by possibility_tac;
result();
(**** Inductive proofs about ns_public ****)
(*can be used to simulate analz_mono_contra_tac
val analz_impI = read_instantiate_sg (sign_of thy)
[("P", "?Y ~: analz (spies ?evs)")] impI;
val spies_Says_analz_contraD =
spies_subset_spies_Says RS analz_mono RS contra_subsetD;
by (rtac analz_impI 2);
by (auto_tac (claset() addSDs [spies_Says_analz_contraD], simpset()));
*)
val parts_induct_tac =
(SELECT_GOAL o EVERY)
[etac reachable.induct 1,
Force_tac 1,
Full_simp_tac 1,
safe_tac (claset() delrules [impI,impCE]),
REPEAT (FIRSTGOAL analz_mono_contra_tac),
ALLGOALS Asm_simp_tac];
(** Theorems of the form X ~: parts (spies evs) imply that NOBODY
sends messages containing X! **)
(*Spy never sees another agent's private key! (unless it's bad at start)*)
(*
Goal "Nprg : Invariant {s. (Key (priK A) : parts (spies s)) = (A : bad)}";
by (rtac InvariantI 1);
by (Force_tac 1);
by (constrains_tac 1);
by Auto_tac;
qed "Spy_see_priK";
*)
Goal "s : reachable Nprg ==> (Key (priK A) : parts (spies s)) = (A : bad)";
by (etac reachable.induct 1);
by (ALLGOALS Force_tac);
qed "Spy_see_priK";
Addsimps [Spy_see_priK];
Goal "s : reachable Nprg ==> (Key (priK A) : analz (spies s)) = (A : bad)";
by Auto_tac;
qed "Spy_analz_priK";
Addsimps [Spy_analz_priK];
AddSDs [Spy_see_priK RSN (2, rev_iffD1),
Spy_analz_priK RSN (2, rev_iffD1)];
(**** Authenticity properties obtained from NS2 ****)
(*It is impossible to re-use a nonce in both NS1 and NS2, provided the nonce
is secret. (Honest users generate fresh nonces.)*)
Goal "[| Crypt (pubK B) {|Nonce NA, Agent A|} : parts (spies s); \
\ Nonce NA ~: analz (spies s); s : reachable Nprg |] \
\ ==> Crypt (pubK C) {|NA', Nonce NA|} ~: parts (spies s)";
by (etac rev_mp 1);
by (etac rev_mp 1);
by (parts_induct_tac 1);
by (ALLGOALS Blast_tac);
qed "no_nonce_NS1_NS2";
(*Adding it to the claset slows down proofs...*)
val nonce_NS1_NS2_E = no_nonce_NS1_NS2 RSN (2, rev_notE);
(*Unicity for NS1: nonce NA identifies agents A and B*)
Goal "[| Nonce NA ~: analz (spies s); s : reachable Nprg |] \
\ ==> EX A' B'. ALL A B. \
\ Crypt (pubK B) {|Nonce NA, Agent A|} : parts (spies s) --> \
\ A=A' & B=B'";
by (etac rev_mp 1);
by (parts_induct_tac 1);
by (ALLGOALS (simp_tac (simpset() addsimps [all_conj_distrib])));
(*NS1*)
by (expand_case_tac "NA = ?y" 2 THEN Blast_tac 2);
(*Fake*)
by (Blast_tac 1);
val lemma = result();
Goal "[| Crypt(pubK B) {|Nonce NA, Agent A|} : parts(spies s); \
\ Crypt(pubK B') {|Nonce NA, Agent A'|} : parts(spies s); \
\ Nonce NA ~: analz (spies s); \
\ s : reachable Nprg |] \
\ ==> A=A' & B=B'";
by (prove_unique_tac lemma 1);
qed "unique_NA";
(*Tactic for proving secrecy theorems*)
val analz_induct_tac =
(SELECT_GOAL o EVERY)
[etac reachable.induct 1,
Force_tac 1,
Full_simp_tac 1,
safe_tac (claset() delrules [impI,impCE]),
ALLGOALS Asm_simp_tac];
(*Secrecy: Spy does not see the nonce sent in msg NS1 if A and B are secure*)
Goal "[| Says A B (Crypt(pubK B) {|Nonce NA, Agent A|}) : set s; \
\ A ~: bad; B ~: bad; s : reachable Nprg |] \
\ ==> Nonce NA ~: analz (spies s)";
by (etac rev_mp 1);
by (analz_induct_tac 1);
(*NS3*)
by (blast_tac (claset() addEs [nonce_NS1_NS2_E]) 4);
(*NS2*)
by (blast_tac (claset() addDs [unique_NA]) 3);
(*NS1*)
by (Blast_tac 2);
(*Fake*)
by (spy_analz_tac 1);
qed "Spy_not_see_NA";
(*Authentication for A: if she receives message 2 and has used NA
to start a run, then B has sent message 2.*)
Goal "[| Says A B (Crypt(pubK B) {|Nonce NA, Agent A|}) : set s; \
\ Says B' A (Crypt(pubK A) {|Nonce NA, Nonce NB|}): set s; \
\ A ~: bad; B ~: bad; s : reachable Nprg |] \
\ ==> Says B A (Crypt(pubK A) {|Nonce NA, Nonce NB|}): set s";
by (etac rev_mp 1);
(*prepare induction over Crypt (pubK A) {|NA,NB|} : parts H*)
by (etac (Says_imp_spies RS parts.Inj RS rev_mp) 1);
by (parts_induct_tac 1);
by (ALLGOALS Clarify_tac);
(*NS2*)
by (blast_tac (claset() addDs [Spy_not_see_NA, unique_NA]) 3);
(*NS1*)
by (Blast_tac 2);
(*Fake*)
by (blast_tac (claset() addDs [Spy_not_see_NA]) 1);
qed "A_trusts_NS2";
(*If the encrypted message appears then it originated with Alice in NS1*)
Goal "[| Crypt (pubK B) {|Nonce NA, Agent A|} : parts (spies s); \
\ Nonce NA ~: analz (spies s); \
\ s : reachable Nprg |] \
\ ==> Says A B (Crypt (pubK B) {|Nonce NA, Agent A|}) : set s";
by (etac rev_mp 1);
by (etac rev_mp 1);
by (parts_induct_tac 1);
by (Blast_tac 1);
qed "B_trusts_NS1";
(**** Authenticity properties obtained from NS2 ****)
(*Unicity for NS2: nonce NB identifies nonce NA and agent A
[proof closely follows that for unique_NA] *)
Goal "[| Nonce NB ~: analz (spies s); s : reachable Nprg |] \
\ ==> EX A' NA'. ALL A NA. \
\ Crypt (pubK A) {|Nonce NA, Nonce NB|} : parts (spies s) \
\ --> A=A' & NA=NA'";
by (etac rev_mp 1);
by (parts_induct_tac 1);
by (ALLGOALS (asm_simp_tac (simpset() addsimps [all_conj_distrib])));
(*NS2*)
by (expand_case_tac "NB = ?y" 2 THEN Blast_tac 2);
(*Fake*)
by (Blast_tac 1);
val lemma = result();
Goal "[| Crypt(pubK A) {|Nonce NA, Nonce NB|} : parts(spies s); \
\ Crypt(pubK A'){|Nonce NA', Nonce NB|} : parts(spies s); \
\ Nonce NB ~: analz (spies s); \
\ s : reachable Nprg |] \
\ ==> A=A' & NA=NA'";
by (prove_unique_tac lemma 1);
qed "unique_NB";
(*NB remains secret PROVIDED Alice never responds with round 3*)
Goal "[| Says B A (Crypt (pubK A) {|Nonce NA, Nonce NB|}) : set s; \
\ ALL C. Says A C (Crypt (pubK C) (Nonce NB)) ~: set s; \
\ A ~: bad; B ~: bad; s : reachable Nprg |] \
\ ==> Nonce NB ~: analz (spies s)";
by (etac rev_mp 1);
by (etac rev_mp 1);
by (analz_induct_tac 1);
by (ALLGOALS (asm_simp_tac (simpset() addsimps [all_conj_distrib])));
by (ALLGOALS Clarify_tac);
(*NS3: because NB determines A*)
by (blast_tac (claset() addDs [unique_NB]) 4);
(*NS2: by freshness and unicity of NB*)
by (blast_tac (claset() addEs [nonce_NS1_NS2_E]) 3);
(*NS1: by freshness*)
by (Blast_tac 2);
(*Fake*)
by (spy_analz_tac 1);
qed "Spy_not_see_NB";
(*Authentication for B: if he receives message 3 and has used NB
in message 2, then A has sent message 3--to somebody....*)
Goal "[| Says B A (Crypt (pubK A) {|Nonce NA, Nonce NB|}) : set s; \
\ Says A' B (Crypt (pubK B) (Nonce NB)): set s; \
\ A ~: bad; B ~: bad; s : reachable Nprg |] \
\ ==> EX C. Says A C (Crypt (pubK C) (Nonce NB)) : set s";
by (etac rev_mp 1);
(*prepare induction over Crypt (pubK B) NB : parts H*)
by (etac (Says_imp_spies RS parts.Inj RS rev_mp) 1);
by (parts_induct_tac 1);
by (ALLGOALS (asm_simp_tac (simpset() addsimps [ex_disj_distrib])));
by (ALLGOALS Clarify_tac);
(*NS3: because NB determines A (this use of unique_NB is more robust) *)
by (blast_tac (claset() addDs [Spy_not_see_NB]
addIs [unique_NB RS conjunct1]) 3);
(*NS1: by freshness*)
by (Blast_tac 2);
(*Fake*)
by (blast_tac (claset() addDs [Spy_not_see_NB]) 1);
qed "B_trusts_NS3";
(*Can we strengthen the secrecy theorem? NO*)
Goal "[| A ~: bad; B ~: bad; s : reachable Nprg |] \
\ ==> Says B A (Crypt (pubK A) {|Nonce NA, Nonce NB|}) : set s \
\ --> Nonce NB ~: analz (spies s)";
by (analz_induct_tac 1);
by (ALLGOALS Clarify_tac);
(*NS2: by freshness and unicity of NB*)
by (blast_tac (claset() addEs [nonce_NS1_NS2_E]) 3);
(*NS1: by freshness*)
by (Blast_tac 2);
(*Fake*)
by (spy_analz_tac 1);
(*NS3: unicity of NB identifies A and NA, but not B*)
by (forw_inst_tac [("A'","A")] (Says_imp_spies RS parts.Inj RS unique_NB) 1
THEN REPEAT (eresolve_tac [asm_rl, Says_imp_spies RS parts.Inj] 1));
by Auto_tac;
by (rename_tac "s B' C" 1);
(*
THIS IS THE ATTACK!
Level 8
!!s. [| A ~: bad; B ~: bad; s : reachable Nprg |]
==> Says B A (Crypt (pubK A) {|Nonce NA, Nonce NB|}) : set s -->
Nonce NB ~: analz (spies s)
1. !!s B' C.
[| A ~: bad; B ~: bad; s : reachable Nprg;
Says A C (Crypt (pubK C) {|Nonce NA, Agent A|}) : set s;
Says B' A (Crypt (pubK A) {|Nonce NA, Nonce NB|}) : set s; C : bad;
Says B A (Crypt (pubK A) {|Nonce NA, Nonce NB|}) : set s;
Nonce NB ~: analz (spies s) |]
==> False
*)