Representation of information

Transcription

1 Representation of information Abstract Syntax Notation No. 1 (ASN.1) and Basic Encoding Rules (BER)

2 ASN.1 Abstract Syntax Notation number 1 is a declarative language used to specify application layer information ASN.1 was originally developed when standardizing protocols for electronic mail in CCITT. Later ASN.1 and its accompanying Basic Encoding Rules where adopted to OSI in order to be used in the presentation layer: standardized data representation. In that way ASN.1 became the general means to specify all application layer information i OSI including management information. Can be compared to XML with regard to information structuring. It s less flexible than XML, but use much more compact encoding

3 Types, Values and Macros By means of ASN.1 it is possible to define types a type is a named set of values There are a number of built-in types in ASN.1 Complex types can be defined using the existing types It is also possible to define values for the types It is also possible to define macros that changes (expands) the grammar of ASN.1 (e.g. like macros do in C/C++)

4 Why ASN.1 is useful It s a protocol language with compilers to produce source code to do encoding and decoding Example: You want to create an application that can do phone number lookup to a server Design the protocol messages Message type (Request? Response?) Data types (phone number, name, address) Specify in ASN.1 syntax Info ::= SEQUENCE { name OCTET STRING, address OCTET STRING, city OCTET STRING, age INTEGER } Compile and link with your application!

5 Example C code Info_t *info; /* Type to encode */ info->age = 35; info->name = George Costanza ; info->city = New York City ; info->address = Broadway 123 ; ber_encode(&asn_def_info, info, sock_write, socket); /* at receiver */ buf = sock_recv(); /* get data from network into buffer */ ber_decode(0, &asn_def_info, (void **)&info, buf, size) Printf( Name: %s, Age: %d\n, info->name, info->age);

6 ASN.1 is a formal language ASN.1 is a language defined in terms of a grammar a grammar is the definition of the syntax of a language using some sort of notation The ASN.1 language defined by a context-free grammar consisting of a number of productions (Backus-Naur Form, BNF) production: name ::= rule (the rule contains terminals and non-terminals) The first production defines the starting symbol ModuleDefinition, which is the biggest syntactic unit of ASN.1

7 A (simplified) grammar of ASN.1 (1) ModuleDefinition ::= Name "DEFINITIONS ::= BEGIN" ModuleBody "END" ModuleBody ::= AssignmentList Empty AssignmentList ::= Assignment AssignmentList Assignment Assignment ::= Name "::=" Type Type ::= ExternalType BuiltinType DefinedType ExternalType ::= Name "." Name BuiltinType ::= PrimitiveType ContructedType TaggedType DefinedType ::= Name PrimitiveType ::= Integer Boolean BitStr OctetStr Any Null ObjId ConstructedType ::= Sequence SequenceOf Set SetOf Choice TaggedType ::= Tag Type Tag "IMPLICIT" Type Integer ::= "INTEGER" "INTEGER" "{" NamedNumberList "}" Boolean ::= "BOOLEAN" BitStr ::= "BIT STRING" "BIT STRING" "{" NamedBitList "}" OctetStr ::= "OCTET STRING" Any ::= "ANY" Null ::= "NULL" ObjId ::= OBJECT IDENTIFIER Sequence ::= "SEQUENCE" "{" ElementTypeList "}" "SEQUENCE {}" SequenceOf ::= "SEQUENCE OF" Type "SEQUENCE"

8 A (simplified) grammar of ASN.1 (2) Set ::= "SET" "{" ElementTypeList "}" "SET {}" SetOf ::= "SET OF" Type "SET" Choice ::= "CHOICE" "{" AlternativeTypeList "}" Tag ::= "[" Class UnsignedNumber "]" Class ::= "UNIVERSAL" "APPLICATION" "PRIVATE" Empty NamedNumberList ::= NamedNumber NamedNumberList "," NamedNumber NamedNumber ::= Name "(" UnsignedNumber ")" Name "(" "-" UnsignedNumber ")" NamedBitList ::= NamedBit NamedBitList "," NamedBit NamedBit ::= Name "(" UnsignedNumber ")" ElementTypeList ::= ElementType ElementTypeList "," ElementType ElementType ::= NamedType NamedType "OPTIONAL" NamedType "DEFAULT" Value NamedType ::= Name Type Type AlternativeTypeList ::= NamedType AlternativeTypeList "," NamedType UnsignedNumber ::= Digit UnsignedNumber Digit Digit ::= "0" "1" "2" "3" "4" "5" "6" "7" "8" "9" Empty ::=

9 Why a grammar? BNF -> Grammar -> Specification -> Implementation With the grammar, we can verify that the specification is syntactically correct... Like a programming language, for (i=0; i<5;i++ ) { }... and it may be interpreted by machines... Like compiling a programming language... and used for information exchange Like executing the compiled program

10 Abstract Syntax for SNMP (1) RFC1157-SNMP DEFINITIONS ::= BEGIN IMPORTS ObjectName, ObjectSyntax, NetworkAddress, IpAddress, TimeTicks FROM RFC1155-SMI; -- top-level message Message ::= SEQUENCE { } version INTEGER { version-1(0) }, community OCTET STRING, data ANY -- contains PDUs HEADER SNMP PDU Version Number Community Name

11 Abstract Syntax for SNMP (2) -- protocol data units PDUs ::= CHOICE { get-request GetRequest-PDU, get-next-request GetNextRequest-PDU, get-response GetResponse-PDU, set-request SetRequest-PDU, trap Trap-PDU } -- PDUs GetRequest-PDU ::= [0] IMPLICIT PDU GetNextRequest-PDU ::= [1] IMPLICIT PDU GetResponse-PDU ::= [2] IMPLICIT PDU SetRequest-PDU ::= [3] IMPLICIT PDU

12 Abstract Syntax for SNMP (3) PDU ::= SEQUENCE { request-id INTEGER, error-status INTEGER { noerror(0), toobig(1), nosuchname(2), badvalue(3), readonly(4), generr(5) }, error-index INTEGER, variable-bindings VarBindList } PDU Type Request ID Error Status Error Index Objects Get/Set/Response header objects to get/set

13 Abstract Syntax for SNMP (4) VarBindList ::= SEQUENCE OF VarBind VarBind ::= -- VarBind for SNMPv2 SEQUENCE { name ObjectName, CHOICE { value ObjectSyntax, unspecified NULL -- the other choices are removed to simplify the example } } END

14 Parse tree Given a string of ASN.1 source code it can be checked that it is correct by drawing a parse tree This is done by applying the productions of the ASN.1 grammar backwards : 0 write the string a the bottom of a paper as one line and encircle all the terminal symbols in it, thus creating a number of leaf nodes in the parse tree to be built 1 then try to identify a contiguous portion of nodes that matches the right-hand side of a production in the grammar 2 make a new node above the identified portion giving it the label of the left-hand side of the production and connect all nodes in the identified portion to it 3 repeat 1 and 2 until there is only one node at the top of the tree

15 Exercise Check that the following Abstract Syntax (ASN.1 source code) is correct according to the ASN.1 grammar. This is a variable binding which is part of an SNMP-PDU: VarBind ::= SEQUENCE { } } name ObjectName, CHOICE { value ObjectSyntax, unspecified NULL -- other choices removed

16 Solution to exercise Assignment Obs! The non-terminal Name in the simplified grammar typeref Type is either a typereference (initial character is upper case) or an BuiltinType identifier (initial character is lower case) ConstructedType Sequence ElementTypeList ElementTypeList ElementType ElementType NamedType NamedType Type BuiltinType identifier Type ConstructedType DefinedType Choice typeref AlternativeTypeList AlternativeTypeList NamedType NamedType identifier Type identifier Type BuiltinType DefinedType PrimitiveType typeref Null VarBind ::= SEQUENCE { name ObjectName, CHOICE { value ObjectSyntax, unspecified NULL } }

17 Basic Encoding Rules (BER) The Basic Encoding Rules is called a Transfer Syntax because it can be used to encode values on bit-level when data is transmitted on the wire. BER is tightly connected to the ASN.1 language every possible value, of every possible type that can be defined using ASN.1, can be encoded with BER Encoding of a particular value contains three (or four) parts the type of the value (often called identifier or tag ) the length of the value the encoding of the value itself (called content ) (if the length of the value is not known then an end-of-contents flag has to be appended to the value)

18 Structure of encoding: Tag-Length-Value Encoding of primitive value where the length is known: Identifier Length Content Encoding of primitive value where the length is not known: Identifier Length octet=80h Content End of content octets= Encoding of a constructed type with two primitive types inside (the length is known): Identifier Length Identifier Length Content Identifier Length Content

19 Encoding of the Identifier (tag/type) octet tag type prim./ constr. tag number 00 UNIVERSAL 01 APPLICATION 10 context-specific 11 PRIVATE 0 primitive type 1 constructed type UNIVERSAL tag numbers: 1 BOOLEAN 2 INTEGER 3 BIT STRING 4 OCTET STRING 5 NULL 6 OBJECT IDENTIFIER 7 OBJECT DESCRIPTOR 8 EXTERNAL 9 REAL 10 ENUMERATED 16 SEQUENCE, SEQUENCE OF 17 SET, SET OF (character string types and time types)

20 Exercise Decode the following ethernet frame containg an SNMPmessage: 00 E0 B0 63 8A D E (IP) AB CD C1 0A DD C4 C1 0A DD C1 (UDP) 04 6E 00 A1 00 3E 67 ED (SNMP) C A F5 1E A 30 0B B B B

21 Solution to exercise (1) 00 E0 B0 63 8A D E destination MAC-addr (Cisco) source MAC-address (3Com) Type=IPv4 prot.type. header len. TOS AB CD C1 0A DD C4 C1 0A DD C1 length=82 frag-id frag.info. TTL Prot=UDP check source= dest= E 00 A1 00 3E 67 ED source-port=1134 dest-port=161(snmp) len=62 check SEQ len=52 INT vers=1 OCT-STR C A F5 1E community= public context 1=get-next-req len=39 INT req-id=62750 INT A 30 0B B err=noerr INT err-ind=0 SEQ len=26 SEQ len=11 OI = { system(1) sysdescr(1) } B B NULL SEQ OI = { system(1) sysobjectid(2) } NULL