The present invention relates to a method for implementing Voice over IP through an electronic device comprising the phase of:

• sampling a voice communication in a corresponding sampled communication;
• encoding the sampled communication in corresponding encoded data;
• decoding the corresponding encoded data in decoded data;

As it is well known, a method implementing Voice over IP transmits a voice communication over a network designed to support a data communication, usually known as packet switched network.

In a packet switched network a plurality of small units of data, usually known as packets, are routed between a sender electronic device and a receiver electronic device on the base of a destination address, stored within each packet.

This type of communication between the sender and the receiver electronic device is known as connectionless communication because each packet may follow a different route on the network, before reaching the receiver device. More particularly, the Internet is based on a connectionless communication as described above.

A method implementing Voice over IP transmits the voice communication between the sender and the receiver electronic devices connected to a first and a second access point of a Packet switched network.

The Internet, that is a Packet switched network used to transmit data communications may be adopted to support voice communications between two electronic devices, for example located in different nations all over the world and connected to a first and a second access point of the Internet.

In fact, the data to be sent are previously divided in a plurality of packets by a sender electronic device, they are sent over the network and they are finally received by a receiver electronic device, wherein they are recomposed in the original data communication.

More particularly, the method implementing Voice over IP must face with the packet loss during the data communication, a consequent bad audio quality of the voice communication being derived from an high packet loss ratio of the corresponding data communication.

In fact, the quality of a voice communication based on a real time voice communication is strictly linked to the percentage of packet loss and to the transmission delay due to the distance between the sender electronic device and the receiver one.

More particularly, experimental results prove that on the Internet the packet loss ratio is between the 0% and the 20%, with delays between 5ms and 500ms. Anyway, when the packet loss rate is over 10% and the delay is over the 150ms the audio quality received at the destination device is considered unacceptable.

A reduction of the packet loss ratio may be achieved through an encoder that reduces the effect of the packet loss on the received data, reconstructing with higher fidelity the corresponding voice communication.

More particularly, a known telephony company developed a method for implementing Voice over IP through a personal computer, at the same time reducing the packet loss ratio.

This method comprises an encoder and a decoder indented to be installed on a sender personal computer for encoding a voice communication into a data communication at sender side and, vice versa, for decoding the data communication into a voice communication at receiver side.

More particularly, the sender personal computer comprises a microphone wherein a first user may speak and a loud speakers wherefrom the first user may hear the voice, for example the voice of a second user in voice communication with him.

The second user is provided with a corresponding receiver personal computer comprising an encoder and a decoder for encoding the voice communication into a data communication at sender side and, vice versa, for decoding the data communication into a voice communication at receiver side. The receiver personal computer also comprises respective microphone and loud speakers.

When the first user speaks, the microphone samples his voice and forwards it to the encoder. The encoder processes the sampled voice and encodes it in a corresponding data communication ready to be sent over the Internet.

The receiver personal computer of the second user receives through the Internet the data communication; the decoder installed on the receiver personal computer of the second user may process and decode the data communication into a corresponding voice communication, ready to be heard from the loud speakers.

Such voice communication is less expensive with respect to a voice communication based on a traditional telephone network, like a circuit switched one, because it has the cost of an internet connection.

Anyway, the use of a personal computer is not always possible to make a telephone call, being the personal computer, for example a notebook, not intended to be used as a mobile phone, especially for its size and weight.

At the same time, an access point to a packet switched Ethernet, like the Internet, is required to implement the voice communication according to the method described above, such access point being not always available.

Moreover, the hardware configuration of a personal computer not always comprises a microphone and a loud dispenser, the installation and use of a headphone as a peripheral device connected to the personal computer being not particularly appreciated by a user.

The problem at the basis of the present invention is that a method for implementing Voice over IP in a packet switched network requires that a personal computer or a notebook, provided with a microphone and a loud dispenser, is connected to an access point of the packet switched network, the use of such personal computer as a mobile telephone being not always possible because the access point to the packet switched network in not ubiquitous and the personal computer is not portable as a mobile phone, especially for its size and weight.

The solution idea on which the present invention is based is that of providing a method for implementing a voice over IP over a packet switched network using a telephone device, more particularly using a mobile phone that is always able to reach the packet switched network and always embed a microphone and a loud dispenser.

This problem is solved, according to the present invention, by a method for implementing Voice over IP as previously indicated and defined by the characterising portion of the enclosed claim 1.

Further characteristics and the advantages of the method implementing Voice over IP according to the present invention will be apparent from the following description of an embodiment thereof, made with reference to the annexed drawings, given for indicative and non-limiting purpose.

• Figure 1a: schematically shows a phase of encoding a sampled communication in corresponding encoded data, according to the present invention.
• Figure 1b: schematically shows a phase of decoding the encoded data in corresponding decoded data, according to the present invention.
• Figure 2a: schematically shows a phase of sampling a voice communication, the phase of encoding the sampled communication in corresponding encoded data and the phase of transmission of the corresponding encoded data, according to the present invention.
• Figure 2b: schematically shows a phase of receiving encoded data, the phase of decoding the encoded data into decoded data and the phase of reproducing a voice communication from decoded data, according to the present invention.
• Figure 3a: schematically shows a phase of sampling a voice communication in a plurality of samples, their grouping in a frame and their encoding, according to the present invention.
• Figure 3b: schematically shows a division of the encoded frames in odd and even samples, their grouping in corresponding packets and the subsequent encryption, according to the present invention.
• Figure 3c: schematically shows at the receiver side, the packet decryption and the reconstruction of the received signal, according to the present invention.
• Figure 4: schematically shows in more detail how the packets of figure 3b are built, comprising odd, even and difference samples, according to the present invention.

With more specific reference to figure 1a, a channel intended to be used for a voice communication is schematically represented and indicated with numeral reference 1.

More particularly, the channel 1 belongs to a packed switched network and the transmission of the voice communication over such network is driven by a method for implementing Voice over IP.

The method provides that an electronic device is connected to the packed switched network, for example to an access point of the channel 1, and samples the voice of a user.

More particularly, the method for implementing Voice over IP comprises:

• a phase for sampling the voice communication of the user in a corresponding sampled communication;
• a phase for encoding the sampled communication in corresponding encoded data;
• a phase for decoding the corresponding encoded data in decoded data;

According to the present invention, the method for implementing Voice over IP comprises a phase for storing an encoder and a decoder inside a portable memory unit of the electronic device, such encoder and decoder being used respectively to execute the phase of encoding the sampled communication and the phase of decoding the corresponding encoded data in decoded data.

More particularly, the encoder and the decoder are stored on a IC Card intended to be inserted in the electronic device.

In figure 1a the transmission of a voice communication from the electronic device over a channel 1 is represented in following block diagrams: the input 2 is a voice communication, sampled for example by a microphone connected or embedded inside the electronic device.

The sampled voice is encoded through an encoder 2, for example stored inside an IC Card included inside the electronic device that is connected to an access point of the channel.

When the sampled communication is encoded, it is transmitted from the electronic device to a second electronic device, for example connected to a second access point of the packed switched network.

The method according to the present invention comprises a phase of decoding the encoded data in a corresponding decoded data. In fact when the second electronic device receives through the packed switched network the encoded data, it decodes them, through the corresponding decoder, in corresponding decoded data. As schematically shown by the diagram block of figure 1b, the received data are decoded through a decoder 5, for example stored inside an IC Card inserted in the receiver electronic device.

The original voice communication of the first user is reproduced by the second electronic device, sending the decoded data to a loud speaker connected or embedded inside the second electronic device.

In figure 2a, it is schematically shown a microphone 6 used to sample the voice communication of a user in a sampled communication and an IC card 7, storing the encoder for encoding the sampled communication in encoded data. The encoded data are transmitted by a mobile phone 8 hosting the IC Card7.

With reference to figure 2b, when a second mobile phone 8 receives the encoded data through the packed switched network, it decodes them through a decoder stored inside an IC Card 7 and forward it to the loud speaker embedded in the mobile phone.

More particularly, the phase of sampling, samples the voice communication through specific mathematic processing, intended to allow the precise reproduction of the voice communication after its transmission over the packet switched network.

As schematically shown in figure 3a, the phase of sampling, samples the voice communication at fs samples/ sec and quantize it at b bit/sample while the phase of encoding groups the quantized samples in frames, compact each frame, divides it in 2 different description and encrypt it before sending.

According to the method of the present invention, the frame is transformed for example by a Discrete Cosine Transform (DCT) in a transformed frame, comprising a plurality of transformed samples.

During the phase of encoding, one or more transformed samples that do not satisfy a condition are discarded. More particularly, those samples that are nearby to zero after the DCT are discarded without any loss in the audio quality. In fact the DCT is an unitary transform and the energy of the signal is preserved after the transformation; so, since after the transformation all the signal is concentrated at low frequency, the high frequency that are nearby to zero may be cut.

The transformed samples are grouped in a plurality of odd transformed samples y1 and in a plurality of even transformed samples y2, as schematically shown in figure 3b.

More particularly, a first difference d1 is processed subtracting the plurality of even transformed samples y2 from the plurality of odd transformed samples y1 and a second difference d2 is processed subtracting the plurality of odd transformed samples y1 from the plurality of even transformed samples y2.

The first difference d1 and the odd transformed samples y1 are packed in a corresponding first packet structure p1, ready to be sent over the packed switched network. Also the second difference d2 and the even transformed samples y2 are packed in a corresponding second packet structure p2.

More particularly, the first packet structure p1 and the second packet structure p2 are encrypted into encrypted packet structures before being transmitted over the packet switched network, for example through a DES encryption.

The encrypted packet structures p1 and p2, represented in major detail in figure 4, are transmitted over the packet switched network from the electronic device; when the second electronic device receives the encrypted packet structures p1 and p2, it decrypts them into a corresponding plurality of packet structures.

The phase of decoding, schematically represented in figure 3c, comprises the decrypting of the encrypted data, for example through a DES^-1 algorithm, and the processing of a validity check for checking the validity of a transmitted packet.

More particularly, the phase of decoding discards, from the corresponding plurality of packets, the first difference d1 and the second difference d2 when said validity check is positive. When the validity check is negative, the phase of decoding uses the first difference d1 and/or the second difference d2 to rebuild one or more of the plurality of packets that are not received correctly by the second electronic device.

With more specific reference to figure 4, an audio signal y is split into two flows: y1, made up of all the odd samples of y, and y2, made up of all the even samples of y. The first difference d1 and the second difference d2 are processed as explained above and two packets p1 and p2 are structured and ciphered before being sent on the network.

At receiver side, the packets p1 and p2 are deciphered and sent to the decoder. More particularly, if p1 and p2 are received correctly, d1 and d2 are discarded and a frame r is built from interleaving of y1 and y2.

If only p1 is correctly received, a frame r1 is built from interleaving of y1 and the difference between y1 and d1.

If only p2 is correctly received, a frame r2 is built from interleaving of the difference between y2 and d2 and y2.

Finally, if both p1 and p2 are not received correctly, a silent frame may be inserted or replaced with white noise or replaced with the last correctly received packet.

Once received the frame r, r1 or r2 a zero padding is performed before the Inverse Discrete Cosine Transformation.

Advantageously, the method implementing Voice Over IP according to the present invention provides that a telephone device, a mobile phone or, more generally, an electronic devices with native embedded microphone and loud speaker is used to transmit and receive a voice communication over a packet switched network.

Advantageously, such electronic device is always connectable to a packet switched network, for example through the same antenna or connection used to reach a circuit switched network.

Advantageously, when a packet switched network in not reachable by the antenna of the electronic device, it may support a voice communication over a reachable circuit switched network.

Advantageously, a telephone device for example a cordless or a mobile phone, being its size and weight appositely designed to be carried, may be always used to reach a packet switched network.

The present invention also relates to an electronic device for implementing Voice over IP.

More particularly, the electronic device comprises:

• sampling means for sampling a voice communication in a corresponding sampled communication;
• encoding means for encoding the sampled communication in corresponding encoded data;
• decoding means for decoding the corresponding encoded data in decoded data;

The electronic device also comprises means to be connected to a packed switched network for transmitting and receiving the corresponding encoded data.

According to the present invention, the electronic device for implementing Voice over IP comprises a portable memory unit for storing the encoding and decoding means.

Advantageously, the portable memory unit is an IC Card intended to be hosted by the electronic device.

Advantageously, the encoding and decoding means stored on the portable memory unit may be uploaded and for example used to implement a Voice communication over IP through another electronic device, able to support a connection with such portable memory device.

The electronic device implementing Voice over IP comprises sampling means for sampling the communication.

Advantageously, the sampling means are embedded inside the electronic device, for example inside a telephone device or a mobile phone.