information frame (HDLC)

The term I-Frame is used in High Level Data Link Control(HDLC) as Information-Frame, but also in MPEG as Intra-Frame. Besides the notation I-Frame there is also the notation IFrame, which stands for Inline-Frame and represents an independent fully functional web page.

1. The I-Frame in HDLC

   In High Level Data Link Control (HDLC), the I-Block or the I-Frame is a standardized transmission format for the transfer of user data as well as for the transmission of control messages in the context of flow control and error recovery. The I- frame consists of the block boundary, followed by the addresses and the control field. This is followed by the data field for user data. The I-frame is the only one of the three frames mentioned with its own data field in which only user data from the higher- level network layer is transported in the form of data blocks. The checksum field and the end flag are then transmitted.
   

   Structure of the HDLC blocks

   The task of the data blocks is to transport data from a data source ofsystem A via a layer-2 connection to a data sink of system B. The data blocks are then transmitted to the data sink of system B via a layer-2 connection. If multiple frames are received, they can be acknowledged individually with an explicit acknowledgement frame, the Receive Ready( RR). The acknowledgement can also be placed in an I-frame sent from system B to system A. Since multiple I-frames may be pending confirmation at the same time on the receive side, they are assigned a transmit sequence number and a receive sequence number to acknowledge the received I-frames from the other side. The maximum number of pending acknowledgements is determined by the window size negotiated between the communication partners.

2. The I-frame in MPEG

   Intra-frame (I-frame) is a frame format of MPEG with which the image area of digital video is divided into macroblocks. The intra-frames are fully stored frames and take into account the image redundancy between the lines. They are formed from luminance and color subsampling and have a size of 8 x 8 pixels. The individual macroblocks contain information about the single image and the temporal relation to the following images. Only every 10th to 15th frame is stored as an intra-frame. Before the intra-frames are compressed, they are transformed into the frequency domain using DCT transformation and then quantized. Compression in JPEG is done using Huffman coding. Intra-frames have file sizes of about 100 kBytes, which are compressed by factors of about 10:1. Predicted frames (P frames) and bidirectional frames are derived from the intra frames. Depending on the film content, intra frames usually appear twice within 30 frames.