Encrypting speech coder

Technical area

The This invention relates generally to communication systems, and used in particular in wireless communication systems, encryption techniques.

General State of the art

Wireless Communication systems use cryptography to provide secure communication means for their subscribers. cryptography provides security so that only an intended recipient the contents of a message that authorized by a transmitter has been sent (and in which it is z. B. to voice data or user data can act FACCH / SACCH messages), can understand and only the authorized transmitter can send the message to the intended recipient. The challenge of cryptography is the content to transform the message into a form that only the intended understand receiver can. This has to happen a way that the intended both for the transmitter and for the receiver is economical. At the same time it must (in terms of time and / or equipment) for a unauthorized recipient (Ie not the intended recipient) be very difficult, to understand the content. As unauthorized receivers and transmitters getting smarter be, the need for secure communications becomes greater.

<figref idrefs="S23">1</figref> shows an encryption speech processor architecture, in a transmitter <figref>10</figref> is integrated on the basis of well-known standards IS-136 (including revisions) of the TIA (Telecommunication Industrial Association) for TDMA (Time Division Multiple Access) and the standard IS-641 for ACELP (Algebraic Code Excited Linear Prediction). The transmitter <figref>10</figref> includes a speech <figref>12</figref>A seven-bit CRC encoder (seven-bit CRC) <figref>14</figref>. a Halbratenfaltungscodierer (½-Rate CC) <figref>16</figref>, dotting <figref>18</figref>. Sprachchiffrierung <figref>20</figref> and a Zweischlitzverschachtler <figref>22</figref>, The speech <figref>12</figref> encodes a message frame, the 160 16-bit speech samples comprising 148 encoded speech bits to to produce 96 class 1 bits and 52 Class 2 bits, the Class 1-bit 48 Class 1A bits and 48 Class 1B bits comprises. The class 1 bits are important Bit (z. B. Bit, the pitch, Intonation, etc. represent) which when transmitted over radio links error control protection require while the error control protection using the CRC code, of the convolutional coding and the bit interleaving is provided. The Class 1A bits as input to 7-Bit CRC <figref>14</figref> supplied to 7 Fehlerkontrollbit to create. The Fehlerkontrollbits, the class-1 bits and 5 tail bits (Convolution code state information include) are used as inputs the ½-Rate CC <figref>16</figref> supplied to to produce 216 code word bits. The code word bits are then an erasure insertion (via puncturing<figref>18</figref>) subjected to produce 208 punctured code word bits.

As next is the Sprachchiffrierung <figref>20</figref> the message saved so that only the intended recipient can understand the content of the message. More specifically, the punctured code word bits and class 2 bits as inputs the Sprachchiffrierung <figref>20</figref> supplied. The Sprachchiffrierung <figref>20</figref> encrypts entries using a 260-bit fixed secret mask associated with the intended recipient is to 260 encrypted bits to create. More specifically, the encryption is by performing an binary XOR operation on the punctured code word bits and the Class 2 bits achieved using the secret mask. The encrypted Bits (by the bit interleaver <figref>22</figref>) Bit interleaved, to produce 260 interleaved bits. Thereafter, the message multiplexed, modulated and transmitted by the transmitter <figref>10</figref> Posted.

The sent message is received by a receiver, not shown, and receiver is the inverse function of the transmitter <figref>10</figref> performed. After receiving the transmitted message, the receiver demodulates the Message sent and demultiplexed to 260 nested Bit to get. The Bitverschachtelungsprozeß then (by a Bitentschachtler) reversed encrypted 260 Bit to get. The encrypted Bits are decrypted (by a Sprachchiffrierer) to a 208 dash codeword bit and 52 Class 2 bits to obtain output. If the receiver by the transmitter <figref>10</figref> used 260-bit fixed secret mask not known (ie the receiver not the intended recipient is) would, recipient not capable of the encrypted Bit properly decipher.

The punctured code word bits are provided to a ½-rate convolutional decoder, in the punctured code word bits are deployed to an output 96 Class 1 bits (including 48 Class 1A bits and 48 Class 1B bits) and 7 to Fehlerkontrollbit receive. Note that in the receiver no inverse function of the erasure insertion process takes place. The bit loss due to erasure insertion is fixed in the development process, as is well known in the art.

The Class 1A bits are of a 7-bit CRC at the receiver to produce a second used amount of 7 Fehlerkontrollbit (wherein the first set of 7 Fehlerkontrollbit part of the output of the ½-rate convolutional are). The&lt;?page 3?&gt;first and second sets are Fehlerkontrollbit (under Using a cyclic redundancy check) compared to determine whether the transmission with respect to the Class 1A bits an error has occurred (ie, determining whether a damaged Framework exists). If no transmission error has occurred, the Class 1 bits and the Class 2 bits (from the Sprachdechiffrierer) for decoding to a speech decoder forwarded. When a transmission error has occurred (ie, a damaged frame is detected), can the Class 1A bits and the 32 most Class 1B bits and discarded by a function or interpolation Class 1A bits and the 32 most Class 1B bits of last undamaged Frame or the last undamaged Frames are replaced and passed to the speech decoder. The Class 2 bits (From the Sprachdechriffierer) and the 16 least significant Class 1B bits (From the ½-rate convolutional decoder) are forwarded to the speech decoder and there together with the forwarded function or interpolation of the Class 1A bits and the 32 most Class 1B bits of the last undamaged frame or the last undamaged Frame decoded. Note that, when bit errors in the Class 2 bits present and the 16 least significant Class 1B bits, such errors a less perceptual impact on speech quality than errors in the other Class 1B and Class 1A bits are having.

The prior art encryption architecture, comprising a Sprachchiffrierung, provides certain advantages. First, a cryptosync from an external source (hereinafter as an "external Cryptosync ") is for synchronizing the 260 bit fixed secret mask at both ends (Ie, in the transmitter and in the receiver) required, cryptosync is a data input, with the assured is that two cryptographic algorithms are synchronized with each other. Secondly there is no degradation in speech quality in the presence of transmission errors. Errors in transmission Class 1A bits and the 32 most significant Class 1B bits can by using a function or interpolation of the Class 1A bits and the 32 most Class 1B bits of the last undamaged frame or the last undamaged frame be masked. The prior art encryption architecture, however, in two respects Security problems vulnerable. First, the 260 bit fixed secret mask, by using determined known plaintext (ie input to the Sprachchiffrierer) are, which can then be used, the encrypted Bit to understand (or decrypt), so that the security of the transmitted Message is compromised. Second, even in the absence of known plaintext simply by XORing adjacent 260-bit frame the hard mask mystery eliminated, and there are information obtained about as the parameters of the ACELP speech algorithm change. consequently a speech processor architecture requires that enhanced security without Using external cryptosync and with minimal speech degradation results.

Out US-A-5,694,473 is a method of decrypting of transmitted again parts a message known, comprising: receiving a message with encryption synchronization and several encrypted Blocks. Until a first block of the plurality of encrypted blocks received with error, is a first keystream from the encryption synchronization and generates at least one of the plurality of encrypted blocks, and using be the first keystream the plurality of encrypted Blocks, received without error before the first block of the plurality encrypted blocks is received with errors, decrypted. is from the message a Primblock determined. If the first block of the plurality of encrypted blocks is received with error, a second message is sent, the retransmission of the first block requests. Upon receiving the retransmission the first block without error, a second keystream from the Primblock and the retransmission of the first block generated, and the decrypting retransmission the first block is using the second keystream decrypted.

Summary the invention

On The method according to the invention is in the independent Claim defined. Preferred forms are set out in dependent claims.

The present invention is an encrypting speech processor architecture, which without using an external cryptosync and with minimal Language deterioration improved safety supplies. this will achieved by integrating a block encryptor and a keystream generator (Instead of a fixed secret mask encryption scheme such. B. the implemented in Sprachchiffrierer) to blocks of encoded speech bits on a transmitter side to encrypt. The block cipher used an invertible cryptographic algorithm and an internal cryptosync to convert a first block of encoded speech bits in a first ciphertext block, the first block of encoded contains speech bit of the first bit plane that significant perceptual would cause degradation of voice quality when while their transmission would occur an error (And falsified Bit instead of a function or interpolation of bits from previous without&lt;?page 4?&gt;damaged would use the framework).

Of the first ciphertext block is then used as input to the keystream generator and a CRC check supplied. Of the Keystream generator used an invertible cryptographic algorithm used as a key-way function is, and an internal cryptosync to implement the first ciphertext block in a key stream, the encrypted first ciphertext blocks comprises. Of the keystream then encryption a second and a third block of encoded speech bits used, a second ciphertext block and a third ciphertext block produce, wherein the third block of encoded speech bits of the bit contains second plane, the slight cause or no perceptual degradation in voice quality when while the transfer of the first ciphertext block, an error occurs, and the second block of encoded speech bits includes bits of the first and the second plane.

at the CRC check is the first ciphertext block to produce Fehlerkontrollbit used, can be recognized with which, if in the transmission the first ciphertext block, an error has occurred, ie, first ciphertext block is a particular type of transmission error detection mechanism protected. The first, second and third ciphertext block and the Fehlerkontrollbit are then processed and transmitted.

On the receiver side is the Fehlerkontrollbit and the received first ciphertext block determines whether during transmission the first ciphertext block, an error has occurred. If no Error is detected, the received first ciphertext block is by a Bloc Kent encrypter decrypts and to decrypt the received second and third ciphertext block (with a the keystream generator used on the transmitter side identical keystream generator). If an error is detected, the first level bits (in the first and the second ciphertext block) using a function or interpolation of the first level bits (in the first and second ciphertext block) using a particular function or interpolation of the first level bits in one or more last undamaged Frame masked. The bit of the second level (in the second and third ciphertext block) can will be directed to a speech decoder, instead decrypted (falsely) to be in order to save processing cycles.

Advantageously, extended this embodiment of the present invention, no error in and out of the bits of the first Level. Specifically extended when an error in any bit in the received first ciphertext block has occurred, such an error never other first level bits because all first level bits by a certain function or interpolation of the first level bits in one or more of the last undamaged frames are masked. Although a at any bit in the received first Ciphertext block error occurring on the bit of the second plane extended, has such an error only little or no perceptual Effect on speech quality. If error in any bit in the received first ciphertext block consist, in particular, no error for any other extended bit of the first level or second level. It is also noted that, due to the keystream generator architecture, wherein the keystream linked XOR with the plaintext is no error in the second or third received extend ciphertext block to the other bits.

Further use of block ciphers and the keystream generator advantageously cryptographic algorithms (or key-way functions), the cryptographically more secure than the 260- bit fixed secret mask are those used in the prior art of Sprachchiffrierern becomes. Thus, the encrypted Speech processor architecture of the present invention cryptographically much safer than the prior art. In addition, internal cryptosync used to be encrypted and decrypted cryptographic algorithms to synchronize. Thus, no external cryptosync required.

Short description tHE dRAWINGS

The Features, aspects and advantages of the present invention are with reference to the following description, appended claims, and appended Drawings. It shows:

<figref idrefs="S23">1</figref> a prior art architecture voice processor, error protection and Sprachchiffrierung;

<figref idrefs="S24">2</figref> a present in a accordance with the invention Transmitter used integrated voice processor architecture;

<figref idrefs="S25">3</figref> a representative Diagram illustrating the coded speech bits, the present application in the be characterized, describes;

<figref idrefs="S26">4</figref> a functional block diagram a block encryptor, the ECMEA algorithms (advanced encryption algorithms for cellular messages) encryption of 48-bit blocks Plaintext to ciphertext according to a Ausfüh&lt;?page 5?&gt;ment of of the present invention;

<figref idrefs="S27">5</figref> a functional block diagram a 48-bit block Kent keyers to perform the reverse function in <figref idrefs="S26">4</figref> imaged block encryptor;

<figref idrefs="S28">6</figref> a functional block diagram a key stream generator, the ECMEA used according to an embodiment the present invention;

<figref idrefs="S29">7</figref> a speech processor architecture, in a present according to the Invention used receiver is integrated;

<figref idrefs="S30">8</figref> a functional block diagram a ECMEA used 40-bit block encryptor according to one embodiment the present invention;

<figref idrefs="S31">9</figref> a functional block diagram a 40-bit block Kent keyers to perform the reverse function in <figref idrefs="S30">8</figref> imaged block encryptor; and

<figref idrefs="S32">10</figref> a functional block diagram a RC5 used 48-bit block encryptor according to one embodiment of the present invention.

Detailed description

FIG. shows in a present according to the Invention used transmitter <figref>30</figref> integrated voice processor architecture. To illustrate the present invention is herein with reference described to a speech processor architecture that well known in the IS-136 standard (including Revisions) for TDMA (Time Division Multiple Access) and the standard IS-641 for ACELP (Algebraic Code Excited Linear Prediction) based. This should, however, on no way limiting of the present invention be construed.

Of the transmitter <figref>30</figref> includes a speech processor architecture with speech <figref>32</figref>. block cipher <figref>34</figref>. Keystream generator <figref>36</figref>, XOR Operators <figref>38</figref>. <figref>40</figref>, Seven-bit CRC encoder (7-Bit CRC) <figref>42</figref>. Halbratenfaltungscodierer (½-Rate CC) <figref>44</figref>. dotting <figref>46</figref> and Zweischlitzverschachteler <figref>48</figref>, The speech processor architecture may be implemented as software, on a computer processor, an application specific integrated executed chip etc. becomes. The speech<figref>32</figref>, The XOR Operators <figref>38</figref>. <figref>40</figref>. the 7-bit CRC <figref>42</figref>, The ½-Rate CC <figref>44</figref>, the dotting <figref>46</figref> and Zweischlitzverschachteler <figref>48</figref> are well known in the art. The block cipher<figref>34</figref> and the Keystream generator <figref>36</figref> will described herein.

Of the speech <figref>32</figref> receives wish to transfer Message frame comprising 160 16-bit speech samples. For purposes of the the present application, the term "speech" and without limitation. a. speech data, user data or control channel messages include. The frame is the speech <figref>32</figref> coded output₃₂Comprising 148 encoded speech bits to produce with coded on first and second-level speech bits, wherein the first level bits represent (encoded speech) voice properties, the voice quality, the will deteriorate noticeable when in any bits of the first Block a transmission error occurs, and the bit of the second level (encoded speech) speech characteristics representing that the voice quality will not deteriorate noticeable when in any bit of the first block a transmission error occurs. Please refer <figref idrefs="S25">3</figref>Wherein a Table 45 is shown, the encoded speech bits which, in the the present application are characterized, describes.

The encoded speech bits are subsequently apportioned and processed in three blocks. The first block (herein as Class 1A bits or output_32-1 will be referred to) contains 48 bits the first level, the second block (the to herein as Class 1B bits or output_32-2 will be referred to) contains 32 bits the first plane and the second plane 16 bits, and the third block (Which herein as Class 2 bits or output_32-3 designated is) contains 52 bits of the second plane. More specifically, with reference to the second block, the first level bits, the most significant bit 32 and the bit the second level, the 16 least significant bits. Note that the present Invention is not limited the number of blocks to which the coded Speech bits are apportioned, or on the number of bits in each Block should be limited. Since the first level bits, the speech quality perceived will deteriorate if a transmission error occurs, error control protection on as many of these bits as possible applied, wherein error control protection z. B. over CRC code and convolution coding provided.

Out the speech <figref>32</figref> be the class 1A bits (first Block) as an input to the block cipher <figref>34</figref>, The Class 1B bits (Second block) as an input to the XOR operator <figref>38</figref> and the Class 2 bits (third block) as input to XOR operator <figref>40</figref> supplied. Of the block cipher <figref>34</figref> used an invertible cryptographic algorithm with the intended recipient associated key to encrypt of plaintext blocks to ciphertext. Such cryptographic algorithms are such. B. RC5 and ECMEA (Enhanced Cellular Messaging Encryption Algorithm). RC5 was developed by Ron Rivest and is in the art &lt;?page 6?&gt;well known. ECMEA was Robert Rance, Daniel Heer, Semyon Mizikovsky et al. developed and on 28.10.1997 at the TIA TR45 Ad-Hoc Authentication Group Meeting disclosed, which was attended by members of the TIA, and is limited and controlled distribution by the TIA considering the export laws of the United States Department of Commerce specified, in the Export Administration Regulations (Titles 15, CFR Parts 730 to including 774) available.

In the block cipher <figref>34</figref> will the Class 1A-bit encrypted, to output₃₄Comprising 48 Class 1A Chiffretextbit (or to produce a first ciphertext block). <figref idrefs="S26">4</figref> shows a functional block diagram of the block encryptor <figref>34</figref>. the ECMEA encryption of 48-bit blocks Plaintext to ciphertext used according to one embodiment of the present invention. The block cipher comprises a first, a second and a third function call <figref>52</figref>. <figref>54</figref> and <figref>58</figref> (Operating in encryption mode to encrypt their inputs) and XOR operator <figref>56</figref>, A first and a the second 16-bit portion of the Class 1A bits (hereinafter referred to as the first or Second Class 1A portion is called) are as details the first and second ECMEA function call <figref>52</figref>. <figref>54</figref> supplied. Of the first EDMEA function call <figref>52</figref> encrypts the first Class 1A portion using the second Class 1A portion as cryptosync input to output₅₂ generating (16 bits).

As in the art is well known, cryptosync is used for synchro- cryptographic encryption and decryption algorithms. In particular ECMEA cryptosync but also serves as a key bit for the purpose of improving the cryptographic strength of ECMEA. In this role, ECMEA is used as a key-way hash function, to both the entropy of the first and second Class 1A portion focus as well as the entire cryptographic strength of 48-bit block encryptor to increase.

output₅₂ is with a third 16 bit portion of the Class 1A bits (hereinafter the is referred to as third-class-1A-part) in the XOR operator 56 XORed, to output₅₆ generating (16 bits).

output₅₆ is as inputs a second and a third ECMEA function call <figref>54</figref>. <figref>58</figref> supplied. Of the second ECMEA function call <figref>54</figref> used output₅₆ as cryptosync to encrypt the first and second Class 1A portion and for generating output_54-1 (16 bit) or output_54-2 (16 bit). The third ECMEA function call <figref>58</figref> encrypted output₅₆ and output_54-1 among Using output_54-2 as cryptosync to output₅₈ generating (32-bit). <figref idrefs="S27">5</figref> shows a functional block diagram of a 48-bit decipherer <figref>35</figref> to implementation the reverse function in <figref idrefs="S26">4</figref> Block encryptor depicted <figref>34</figref>,

output₅₈ and output_54-2 include together output₃₄ the block encryptor <figref>34</figref>, output₃₄ is as inputs the 7-bit CRC <figref>42</figref>. the ½-Rate CC <figref>44</figref> and the keystream generator <figref>36</figref> supplied. In the 7-bit CRC <figref>42</figref> is with output₃₄ the 7 Fehlerkontrollbit comprehensive output₄₂ generated. output₄₂ thereafter as input to ½-Rate CC <figref>44</figref> supplied.

Of the Keystream generator <figref>36</figref> used either an invertible cryptographic algorithm, such. as RC5 and ECMEA, or a suitable non-invertible key-way function, wherein the intended recipient a key to Issue a Bitschlüsselstroms assigned. Note that a One-way function is a general type of cryptographic operation is and actually invertible cryptographic algorithms includes a subclass. A Who observes the output of a key-way function can not on the key yet close to the input.

Of the Keystream generator <figref>36</figref> becomes by the 48 Class 1A Chiffretextbit (ie output₃₄) driven at the input of the key stream generator <figref>34</figref> often are repeated enough to fill a 104-bit block. In the keystream generator <figref>36</figref> will Class 1A Chiffretextbit encrypted output₃₆ of Keystream, comprising 100 bit encrypted to produce class-1A Chiffretextbit. More specifically, output₃₆ generated by the (of the key stream generator <figref>36</figref> used) cryptographic algorithm by encrypting publicly known pseudo random data accomplished is where it is actually located is the class-1A Chiffretextbit. Note that 4 of the Class 1A Chiffretextbit (which include the 104-bit block) discarded be because ECMEA (and most other block algorithms) only encrypt a whole number of bytes.

<figref idrefs="S28">6</figref> shows a functional block diagram of the keystream generator <figref>36</figref>. the ECMEA used according to an embodiment of the present invention. The key stream generator<figref>36</figref> includes the ECMEA function call <figref>62</figref>. of the 104-bit block input comprising repeated output₃₄ (From the block cipher <figref>34</figref>), Receives. one Note that the Output of Block Encryptor <figref>34</figref> at attached himself or is concatenated with itself, in order to obtain 96-bit, and then partially appended one more time is to obtain a 104-bit block input to Keystream Generator. The 104-bit block input is the ECMEA function call <figref>62</figref> among Using the 16 most Bit of output₃₄ (Ie &lt;?page 7?&gt;16 most Bit the 48 Class 1A ciphertext bits) encrypted as cryptosync to output₃₆ to create.

output₃₆ is as inputs to XOR Operators <figref>38</figref> and <figref>40</figref> supplied. More accurate specifically, a 48-bit part is used by output₃₆ as Input to the XOR operator <figref>38</figref> and a 52-bit portion of output₃₆ as input to XOR operator <figref>40</figref> supplied. In the XOR operator <figref>38</figref> be the class 1B bits (ie, the second block) with the 48 bit portion of output₃₆ XORed to output <figref>38</figref>Comprising 48 Class 1B Chiffretextbit (ie to produce the second ciphertext block). output₃₈ becomes then as input to ½-Rate CC <figref>44</figref> supplied. Similarly in the XOR operator <figref>40</figref> the Class 2 bits (ie, the third block) with the 52 bit portion of output₃₆ XORed to output₄₀Comprising 52 Class 2 Chiffretextblockbit (Ie, the third ciphertext block) to produce. output₄₀ is then used as input to the Zweischlitzverschachteler <figref>48</figref> supplied.

In the ½-Rate CC <figref>44</figref> will output₄₂, output₃₄ and output₃₈ along with 5 tail bits (the Convolution code state information includes) for generating output₄₄ (Comprising 216 code word bits) used. output₄₄ is the puncturing <figref>46</figref> supplied to output₄₆Comprising 208 punctured code word bits, to create. output₄₆ is then the Zweischlitzverschachteler <figref>48</figref> supplied and there with output₄₀ bit interleaved to output₄₈Comprising 260 interleaved bits to produce. output₄₈ is then multiplexed, modulated and transmitted by the transmitter <figref>30</figref> Posted.

The sent message is received by a receiver in which the inverse function of the transmitter <figref>10</figref> is carried out. <figref idrefs="S29">7</figref> showing a speech processor architecture, in a present according to the Invention used receiver <figref>70</figref> integrated is. Recipient<figref>70</figref> includes a Speech processor architecture with the speech decoder <figref>72</figref>. the Bloc Kent encrypter <figref>74</figref>, the Keystream generator <figref>76</figref>. XOR Operators <figref>78</figref>. <figref>80</figref>, The 7-bit CRC <figref>82</figref>. the ½-rate convolutional decoder <figref>84</figref> and the Zweischlitzbitentschachteler <figref>86</figref>, The speech processor architecture may be implemented as software on a computer processor is running an application specific integrated chip, etc.. The Spachdecodierer <figref>72</figref>, The XOR Operators <figref>78</figref>. <figref>80</figref>. the 7-bit CRC <figref>82</figref>, The ½-rate convolutional decoder <figref>84</figref> and the Zweischlitzbitentschachtler <figref>86</figref> are all in the art well known. The block cipher<figref>74</figref> and the keystream generator <figref>76</figref> will described herein.

After receiving the transmitted message, the receiver performs a Demodulating and demultiplexing the transmitted message through to output_{48 '} (D. h. 260 to obtain interleaved bits), wherein the Apostrophnotation mean a received version of the corresponding transmitted version should. The Zweischlitzbitentschachtler<figref>86</figref> used output_{48 '} to Generating output_{46 '} (Ie, 208 punctured code word bits and 52 class 2 Chiffretextbit or the third ciphertext block). The 208 punctured code word bits are then passed through the ½-rate convolutional decoder <figref>84</figref> unfolded 5 tail bits, output_{42 '} (Ie 7 Fehlerkontrollbit) output_{34 '} (D. h. 48 Class 1A Chiffretextbit or the first ciphertext block) and output_{38 '} (D. h. 48 class-1B Chiffretextbit or to obtain the second ciphertext block). Note that in the receiver <figref>80</figref> no Operation for reversing the erasure insertion process puncturing <figref>46</figref> takes place. The lost due to deletion insertion Bits in the receiver through the ½-rate convolutional decoder <figref>84</figref> restored.

output_{34 '} becomes as input to 7-Bit CRC <figref>82</figref> supplied to a second set of to produce 7 Fehlerkontrollbit (the first set of 7 Fehlerkontrollbit together output_{42 '} includes what the 7 Fehlerkontrollbit are represented by the ½-rate convolutional decoder be issued). The first and the second set of Fehlerkontrollbit are examined to determine whether an error in the transmission the first ciphertext block (ie, the Class 1A Chiffretextbit) was present (ie determining whether a damaged frame exists). If no transmission error has occurred is output_{34 '} (Ie class-1A Chiffretextbit) the Bloc Kent encrypter <figref>74</figref> and the keystream generator <figref>76</figref> forwarded. Of the Bloc Kent encrypter is operable to the inverse function of Block Encryptor <figref>34</figref> perform.

Please refer <figref idrefs="S27">5</figref>Wherein a functional block diagram of a 48-bit decipherer <figref>35</figref> to implementation the reverse function in <figref idrefs="S26">4</figref> pictured Block Encryptor <figref>34</figref> ready is. The key stream generator<figref>76</figref> is with the key stream generator <figref>36</figref> identical.

In the Bloc Kent encrypter <figref>74</figref> becomes output_{34 '} decrypted to output_{32-1 '} (D. h. to obtain 48 Class 1A bits). In the key stream generator<figref>76</figref> will with output_{34 '} output_{36 '} of keystream (Ie, 100 encrypted Class 1A Chiffretextbit) generated. The key stream is then output_{38 '} (D. h. Class 1B Chiffretextbit) and output_{40 '} (Ie class 2 Chiffretextbit) XORed to output_{32-2 '} (ie Class 1B bits) and output_{32-3 '} (Ie, Class 2 bits) to be obtained. output_{32-1 '}. output_{32-2 '} and output_{32-3 '} (D. h. Class 1A bits, Class 1B bits and Class 2 bits) are determined by the Speech decoder <figref>72</figref> decoded to the original receive message frames.

If a transmission error has occurred, the first level bits (ie, Klas&lt;?page 8?&gt;se-1A-bit and the 32 most Class 1B bits) and discarded with a function or interpolation of First level bits of one or more previous frames undamaged replaced (as in the well-known IS-641 described) and to the speech decoder <figref>72</figref> forwarded. A block decryption Class 1A bits or generation of a key stream (Using the key stream generator <figref>76</figref>) would not be necessary, since this data irrelevant in the case of a failing CRC would. Thus, it is not possible, (Ie, the Class 2 bits and the 16 least significant bit of the second plane Class 1B bits) to decrypt. Avoiding these tasks, saving processor cycles. The deterioration the voice quality due Encrypts the staying this bit of the second level is little or no perceptual have impact.

one Note that the present invention in a similar Way as voice confidentiality FACCH and SACCH encryption provides, where is the origin of cryptosync instead external internal. The first word each FACCH and SACCH message includes an 8-bit message type unsecured must remain. Thus would the first word of each FACCH and SACCH message includes a 40-bit block cipher <figref>80</figref> (and -entschlüsseler <figref>90</figref>) use, such as in <figref idrefs="S30">8</figref> (and in <figref idrefs="S31">9</figref>) Shown, and the remaining words would be the 48-bit block cipher of <figref idrefs="S26">4</figref> use.

Although The present invention in great detail with reference to certain described embodiments was, other versions are possible. To the Example are other encryptor and descrambler possible. Please refer <figref idrefs="S32">10</figref>Wherein a 48-bit block cipher <figref>100</figref>. is used the RC5, ready. For those of ordinary skill seen that the Concept of the present invention to speech processor architectures can be applied to other types of speech, error control and encryption and / or using cryptographic algorithms.